---
title: Action Gateway
description: Action Gateway gives AI agents governed access to tools over the Model Context Protocol (MCP). The `/v2/action-gateway` endpoints manage the tool …
product: Reference
url: https://docs.digitalocean.com/reference/api/reference/action-gateway/
---

> **For AI agents:** The documentation index is at [https://docs.digitalocean.com/llms.txt](https://docs.digitalocean.com/llms.txt). Markdown versions of pages use the same URL with `index.html.md` in place of the HTML page (for example, append `index.html.md` to the directory path instead of opening the HTML document).

# Action Gateway

Generated on 8 Oct 2026

Action Gateway gives AI agents governed access to tools over the Model Context Protocol (MCP). The `/v2/action-gateway` endpoints manage the tool catalog and the resources that control how agents use it:

- **Tools** browse and search the catalog, its providers, and global provider and tool health.
- **Toolbelts** are versioned, team-owned sets of tools that sessions and policies reference as `toolbelt:<name>@<version>`.
- **Output views** reshape a tool's output before it reaches the agent.
- **MCP servers** register your own remote MCP servers and choose which of their tools to enable.
- **Connections** authorize one of your end users to use a provider's tools, for example through OAuth.
- **Sessions** bind a tool selection and permission policy to an MCP URL that an agent connects to.
- **Users** and **end user limits** show and limit activity per end user, identified by a session's `actorId`.

Base URL `https://api.digitalocean.com`

## Endpoints

- **GET** [Get an Actor's Limits](#actorLimits_get)
- **POST** [Clear an Actor's Limits](#actorLimits_delete)
- **POST** [Set an Actor's Limits](#actorLimits_update)
- **GET** [List Connections](#connections_list)
- **POST** [Create a Connection](#connections_create)
- **GET** [Get a Connection](#connections_get)
- **DELETE** [Delete a Connection](#connections_delete)
- **GET** [List MCP Servers](#mcpServers_list)
- **POST** [Register an MCP Server](#mcpServers_create)
- **GET** [Get an MCP Server](#mcpServers_get)
- **PATCH** [Update an MCP Server](#mcpServers_patch)
- **DELETE** [Delete an MCP Server](#mcpServers_delete)
- **POST** [Discover an MCP Server's Tools](#mcpServers_run_resync)
- **GET** [List an MCP Server's Tools](#mcpServers_list_tools)
- **PUT** [Set Which of an MCP Server's Tools Are Enabled](#mcpServers_update_tools)
- **GET** [List Output Views](#outputViews_list)
- **POST** [Create an Output View](#outputViews_create)
- **POST** [Preview an Output View](#outputViews_validate)
- **GET** [Get an Output View](#outputViews_get)
- **DELETE** [Delete an Output View](#outputViews_delete)
- **GET** [List Sessions](#sessions_list)
- **POST** [Create a Session](#sessions_create)
- **GET** [Search Sessions](#sessions_list_search)
- **DELETE** [Delete a Session](#sessions_delete)
- **GET** [List Toolbelts](#toolbelts_list)
- **POST** [Create a Toolbelt](#toolbelts_create)
- **GET** [Search Toolbelts](#toolbelts_list_search)
- **GET** [Get a Toolbelt](#toolbelts_get)
- **DELETE** [Delete a Toolbelt](#toolbelts_delete)
- **GET** [List the Providers in a Toolbelt](#toolbelts_list_providers)
- **GET** [List a Toolbelt's Tools from One Provider](#toolbelts_list_provider_tools)
- **POST** [Add Tools to a Toolbelt](#toolbelts_add_tools)
- **POST** [Remove Tools from a Toolbelt](#toolbelts_delete_tools)
- **GET** [List Tools](#tools_list)
- **GET** [List Tool Health](#tools_list_health)
- **GET** [List Provider Health](#tools_list_health_providers)
- **GET** [Get Tool Health](#tools_get_health)
- **GET** [List Providers](#tools_list_providers)
- **GET** [Search Providers](#tools_list_search_providers)
- **GET** [Search Tools](#tools_list_search)
- **GET** [List Toolkits](#tools_list_toolkits)
- **GET** [List Users](#users_list)
- **GET** [Get a User](#users_get)

## GET Get an Actor's Limits

`/v2/action-gateway/actors/{actor_id}/limits`

**Authorizations: bearer_auth (1 scope)**

Http: Bearer

Required scopes: action_gateway:read

### OAuth Authentication

In order to interact with the DigitalOcean API, you or your application must authenticate.

The DigitalOcean API handles this through OAuth, an open standard for authorization. OAuth allows you to delegate access to your account. Scopes can be used to grant full access, read-only access, or access to a specific set of endpoints.

You can generate an OAuth token by visiting the [Apps & API](https://cloud.digitalocean.com/account/api/tokens) section of the DigitalOcean control panel for your account.

An OAuth token functions as a complete authentication request. In effect, it acts as a substitute for a username and password pair.

Because of this, it is absolutely **essential** that you keep your OAuth tokens secure. In fact, upon generation, the web interface will only display each token a single time in order to prevent the token from being compromised.

DigitalOcean access tokens begin with an identifiable prefix in order to distinguish them from other similar tokens.

- `dop_v1_` for personal access tokens generated in the control panel
- `doo_v1_` for tokens generated by applications using [the OAuth flow](https://docs.digitalocean.com/reference/api/oauth-api/index.html.md)
- `dor_v1_` for OAuth refresh tokens

#### Scopes

Scopes act like permissions assigned to an API token. These permissions determine what actions the token can perform. You can create API tokens that grant read-only access, full access, or limited access to specific endpoints by using custom scopes.

Generally, scopes are designed to match HTTP verbs and common CRUD operations (Create, Read, Update, Delete).

| HTTP Verb | CRUD Operation | Scope |
|---|---|---|
| GET | Read | `<resource>:read` |
| POST | Create | `<resource>:create` |
| PUT/PATCH | Update | `<resource>:update` |
| DELETE | Delete | `<resource>:delete` |

For example, creating a new Droplet by making a `POST` request to the `/v2/droplets` endpoint requires the `droplet:create` scope while listing Droplets by making a `GET` request to the `/v2/droplets` endpoint requires the `droplet:read` scope.

Each endpoint below specifies which scope is required to access it when using custom scopes.

#### How to Authenticate with OAuth

In order to make an authenticated request, include a bearer-type `Authorization` header containing your OAuth token. All requests must be made over HTTPS.

#### Authenticate with a Bearer Authorization Header

```
curl -X $HTTP_METHOD -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" "https://api.digitalocean.com/v2/$OBJECT"
```

Returns the per-minute limits set for the actor and the limits in effect, each the lower of your team's limit and the actor's override. Returns 404 when the actor has no sessions, connections, or limit overrides.

#### Path Parameters

`actor_id` string required

Example: `alice`

Actor ID: a session `actor_id` or a connection `user_id`, 1 to 64 characters from `[A-Za-z0-9._-]`.

##### Request: `/v2/action-gateway/actors/{actor_id}/limits`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/actors/alice/limits"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`configured_limits` array of object optional

Overrides set for the actor; categories without one are omitted.

**Show child properties**

`category` string required

The category the limit applies to.

`requests_per_minute` string (uint64) required

Example: `30`

Calls allowed per minute. 0 blocks every call in the category.

`effective_limits` object optional

Limits in effect for the actor in every category: the lower of your team's limit and the actor's override.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "configured_limits": [
    {
      "category": "LIMIT_CATEGORY_WEB_SEARCH_REQUESTS_PER_MINUTE",
      "requests_per_minute": "10"
    }
  ],
  "effective_limits": {
    "native_tool_calls_requests_per_minute": "600",
    "non_native_tool_calls_requests_per_minute": "600",
    "web_fetch_requests_per_minute": "120",
    "web_search_requests_per_minute": "10"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Clear an Actor's Limits

`/v2/action-gateway/actors/{actor_id}/limits:clear`

Removes the actor's overrides for the listed categories, which then follow your team's limits again. Clearing a category without an override succeeds.

#### Path Parameters

`actor_id` string required

Example: `alice`

Actor ID: a session `actor_id` or a connection `user_id`, 1 to 64 characters from `[A-Za-z0-9._-]`.

#### Request Body: `application/json`

`categories` array of string, one of: LIMIT_CATEGORY_WEB_SEARCH_REQUESTS_PER_MINUTE, LIMIT_CATEGORY_WEB_FETCH_REQUESTS_PER_MINUTE, LIMIT_CATEGORY_NATIVE_TOOL_CALLS_REQUESTS_PER_MINUTE, LIMIT_CATEGORY_NON_NATIVE_TOOL_CALLS_REQUESTS_PER_MINUTE required

Example: `["LIMIT_CATEGORY_WEB_SEARCH_REQUESTS_PER_MINUTE"]`

Required. Categories to clear, each at most once.

##### Request: `/v2/action-gateway/actors/{actor_id}/limits:clear`

### Payload

Content type `application/json`

```json
{
  "categories": [
    "LIMIT_CATEGORY_WEB_SEARCH_REQUESTS_PER_MINUTE"
  ]
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"categories":["LIMIT_CATEGORY_WEB_SEARCH_REQUESTS_PER_MINUTE"]}' \
  "https://api.digitalocean.com/v2/action-gateway/actors/alice/limits:clear"
```

#### Responses

**200** The overrides were removed.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Set an Actor's Limits

`/v2/action-gateway/actors/{actor_id}/limits:set`

Sets per-minute limits for the listed categories and leaves the others unchanged. A limit may not exceed your team's current limit for the category (400 otherwise), and 0 blocks every call in the category. The actor does not need to exist yet.

#### Path Parameters

`actor_id` string required

Example: `alice`

Actor ID: a session `actor_id` or a connection `user_id`, 1 to 64 characters from `[A-Za-z0-9._-]`.

#### Request Body: `application/json`

`overrides` array of object required

Required. Limits to set, at most one per category.

**Show child properties**

`category` string required

The category the limit applies to.

`requests_per_minute` string (uint64) required

Example: `30`

Calls allowed per minute. 0 blocks every call in the category.

##### Request: `/v2/action-gateway/actors/{actor_id}/limits:set`

### Payload

Content type `application/json`

```json
{
  "overrides": [
    {
      "category": "LIMIT_CATEGORY_WEB_SEARCH_REQUESTS_PER_MINUTE",
      "requests_per_minute": "10"
    }
  ]
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"overrides":[{"category":"LIMIT_CATEGORY_WEB_SEARCH_REQUESTS_PER_MINUTE","requests_per_minute":"10"}]}' \
  "https://api.digitalocean.com/v2/action-gateway/actors/alice/limits:set"
```

#### Responses

**200** The overrides were set.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Connections

`/v2/action-gateway/connections`

Returns the connections you can see, excluding revoked ones. Offset-paged with page and `per_page`; filter by provider, `user_id`, and status.

#### Query Parameters

`provider` string optional

Example: `jira`

Optional provider slug, matched exactly.

`user_id` string optional

Example: `alice`

Optional substring of `user_id`. It must use the `user_id` alphabet, `[A-Za-z0-9._-]`.

`status` string, one of: pending, active, expired optional

Example: `active`

pending|active|revoked|expired. List results always omit revoked connections, including when this filter is revoked.

`sort` string, one of: created_at, provider, user_id, status optional

Example: `created_at`

`created_at`, provider, `user_id`, or status. Defaults to provider, then `user_id`.

`sort_direction` string, one of: asc, desc optional

Example: `desc`

asc (the default) or desc.

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

##### Request: `/v2/action-gateway/connections`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/connections?provider=jira"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`connections` array of object optional

Connections on this page.

**Show child properties**

`api_key` object optional

Set for `team_api_key` connections.

`connection_parameters` object optional Nullable

Example: `{"site_url":"https://example.atlassian.net"}`

Non-sensitive provider configuration.

`created_at` string (date-time) optional Nullable

Example: `2026-07-14T00:50:00Z`

When the connection was created.

`credential_id` string optional

Empty for `digitalocean_oauth`; otherwise the ID of the team provider credential the connection uses.

`credential_kind` string, one of: digitalocean_oauth, private_oauth, team_api_key optional

Example: `digitalocean_oauth`

`digitalocean_oauth`, `private_oauth`, or `team_api_key`.

`granted_at` string (date-time) optional Deprecated Nullable

Example: `2026-07-14T00:52:10Z`

Deprecated: read `oauth.granted_at`.

`id` string (uuid) optional

Example: `39f9f768-a7f9-4c91-9557-3328c53dd054`

Opaque connection ID.

`network` object optional

Private network the connection's calls use, when configured.

`oauth` object optional

Set for `digitalocean_oauth` and `private_oauth` connections.

**Show child properties**

`granted_at` string (date-time) optional Nullable

Example: `2026-07-14T00:52:10Z`

When the user completed authorization.

`scopes` array of string optional

Example: `["write:jira-work"]`

Provider scopes granted, or requested while pending.

`owning_user_id` string optional

Example: `2f8e1c3a-6b4d-4f7e-9a1c-5d3b7e9f1a2c`

DigitalOcean user ID of the user who created the connection, when recorded.

`provider` string optional

Example: `jira`

Provider slug, for example `jira`.

`provider_display_name` string optional

Example: `Jira`

Human-readable provider name, for example `Jira`.

`revoked_at` string (date-time) optional Nullable

When the connection was revoked.

`scopes` array of string optional Deprecated

Example: `["write:jira-work"]`

Deprecated: read `oauth.scopes`. Duplicated here so a reader of the pre-union shape keeps working; empty for a kind with no OAuth grant.

`status` string, one of: pending, active, revoked, expired optional

Example: `active`

pending, active, revoked, or expired.

`updated_at` string (date-time) optional Nullable

Example: `2026-07-14T00:52:10Z`

When the connection was last modified.

`user_id` string optional

Example: `alice`

Your identifier for the user the connection acts for.

`pagination` object optional

Paging applied to this response and the total number of connections.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "connections": [
    {
      "connection_parameters": {
        "site_url": "https://example.atlassian.net"
      },
      "created_at": "2026-07-14T00:50:00Z",
      "credential_id": "",
      "credential_kind": "digitalocean_oauth",
      "granted_at": "2026-07-14T00:52:10Z",
      "id": "39f9f768-a7f9-4c91-9557-3328c53dd054",
      "network": null,
      "oauth": {
        "granted_at": "2026-07-14T00:52:10Z",
        "scopes": [
          "write:jira-work"
        ]
      },
      "provider": "jira",
      "provider_display_name": "Jira",
      "revoked_at": null,
      "scopes": [
        "write:jira-work"
      ],
      "status": "active",
      "updated_at": "2026-07-14T00:52:10Z",
      "user_id": "alice"
    }
  ],
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 1
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Create a Connection

`/v2/action-gateway/connections`

Connects one of your users, identified by `user_id`, to a provider so tools can act on their account. An OAuth connection starts pending, and authorization carries the URL to send the user to; read the connection until it is active or expired. A connection through a team API-key credential is active immediately.

There is one connection per provider and `user_id` for each DigitalOcean user who creates connections. Repeating the request returns the existing connection when its scopes cover the request, reusing a pending connection's link, and restarts an expired or revoked connection under the same ID. Requesting scopes beyond an existing connection's, or a different credential, returns 400 until that connection is deleted.

#### Request Body: `application/json`

`connection_parameters` object optional

Example: `{"site_url":"https://example.atlassian.net"}`

Values for the provider's `connection_parameters`, validated against their specifications.

`credential` object optional

Optional credential to connect through. Omitted uses DigitalOcean's shared OAuth application.

**Show child properties**

`digitalocean_oauth` object optional

Use DigitalOcean's shared OAuth application.

`team_credential` object optional

Use one of your team's provider credentials.

**Show child properties**

`credential_id` string optional

Example: `8d6f2b1a-3c4e-4d5f-9a8b-7c6d5e4f3a2b`

Required. ID of an active provider credential for the same provider that you can use.

`network` object optional

Optional private network for the connection's calls. Requires VPC networking to be enabled for your account (403 otherwise).

**Show child properties**

`vpc` object optional

Route through a VPC.

**Show child properties**

`destinations` array of object optional

Private hosts the connection's calls may reach through the VPC. At least one is required.

**Show child properties**

`allowed_ip_cidrs` array of string optional

Example: `["10.116.0.0/20"]`

Optional CIDR ranges; when set, every address the host resolves to must fall within one of them.

`host` string optional

Example: `jira.internal.example.com`

Host name or IP address. It must resolve only to private addresses.

`port` integer (uint32) optional

Example: `443`

Port, 1 to 65535.

`vpc_uuid` string (uuid) optional

Example: `5a4981aa-9653-4bd1-bef5-d6bff52042e4`

UUID of a VPC owned by your team.

`provider` string required

Example: `jira`

Required provider slug, from the provider list.

`scopes` array of string optional

Example: `["write:jira-work"]`

Optional OAuth scopes to request. Defaults to every scope the provider offers; scopes outside that set are rejected. Ignored for API-key credentials.

`user_id` string required

Example: `alice`

Required. Your identifier for the user the connection acts for: 1 to 64 characters from `[A-Za-z0-9._-]`. A session whose `actor_id` equals it uses this connection.

##### Request: `/v2/action-gateway/connections`

### Payload

Content type `application/json`

```json
{
  "connection_parameters": {
    "site_url": "https://example.atlassian.net"
  },
  "credential": {
    "digitalocean_oauth": {}
  },
  "provider": "jira",
  "scopes": [
    "write:jira-work"
  ],
  "user_id": "alice"
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"provider":"jira","user_id":"alice","scopes":["write:jira-work"],"connection_parameters":{"site_url":"https://example.atlassian.net"},"credential":{"digitalocean_oauth":{}}}' \
  "https://api.digitalocean.com/v2/action-gateway/connections"
```

#### Responses

**200** The connection was created.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`authorization` object optional

Present only while the connection is pending.

`connection` object optional

The connection.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**409** The request could not be completed due to a conflict.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "authorization": {
    "connect_url": "https://cloud.digitalocean.com/security/connectlinks/confirm?token=eyJhbGciOi",
    "expires_at": "2026-07-14T01:00:00Z",
    "status": "requires_authorization",
    "verification_code": "97rwxzcy"
  },
  "connection": {
    "connection_parameters": {
      "site_url": "https://example.atlassian.net"
    },
    "created_at": "2026-07-14T00:50:00Z",
    "credential_id": "",
    "credential_kind": "digitalocean_oauth",
    "granted_at": null,
    "id": "39f9f768-a7f9-4c91-9557-3328c53dd054",
    "network": null,
    "oauth": {
      "granted_at": null,
      "scopes": [
        "write:jira-work"
      ]
    },
    "provider": "jira",
    "provider_display_name": "Jira",
    "revoked_at": null,
    "scopes": [
      "write:jira-work"
    ],
    "status": "pending",
    "updated_at": "2026-07-14T00:50:00Z",
    "user_id": "alice"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**409**

```json
{
  "id": "conflict",
  "message": "The request could not be completed due to a conflict."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Get a Connection

`/v2/action-gateway/connections/{id}`

Returns the connection and, while it is pending, the authorization the user still has to complete. Reading a pending connection picks up a completed authorization, and marks the connection expired once its link has lapsed. Returns 404 for an unknown connection or one you cannot see.

#### Path Parameters

`id` string required

Example: `39f9f768-a7f9-4c91-9557-3328c53dd054`

Connection ID.

##### Request: `/v2/action-gateway/connections/{id}`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/connections/39f9f768-a7f9-4c91-9557-3328c53dd054"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`authorization` object optional

Present only while the connection is pending and you created it.

`connection` object optional

The connection.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "authorization": null,
  "connection": {
    "connection_parameters": {
      "site_url": "https://example.atlassian.net"
    },
    "created_at": "2026-07-14T00:50:00Z",
    "credential_id": "",
    "credential_kind": "digitalocean_oauth",
    "granted_at": "2026-07-14T00:52:10Z",
    "id": "39f9f768-a7f9-4c91-9557-3328c53dd054",
    "network": null,
    "oauth": {
      "granted_at": "2026-07-14T00:52:10Z",
      "scopes": [
        "write:jira-work"
      ]
    },
    "provider": "jira",
    "provider_display_name": "Jira",
    "revoked_at": null,
    "scopes": [
      "write:jira-work"
    ],
    "status": "active",
    "updated_at": "2026-07-14T00:52:10Z",
    "user_id": "alice"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## DELETE Delete a Connection

`/v2/action-gateway/connections/{id}`

Marks the connection revoked and returns it; for an OAuth connection, its stored tokens are deleted first. Creating a connection for the same provider and `user_id` later reuses the same ID. Only the user who created the connection can delete it; for anyone else it is 404.

#### Path Parameters

`id` string required

Example: `39f9f768-a7f9-4c91-9557-3328c53dd054`

Connection ID.

##### Request: `/v2/action-gateway/connections/{id}`

### cURL

```bash
curl -X DELETE \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/connections/39f9f768-a7f9-4c91-9557-3328c53dd054"
```

#### Responses

**200** The connection was revoked.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`connection` object optional

The resulting revoked connection.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "connection": {
    "connection_parameters": {
      "site_url": "https://example.atlassian.net"
    },
    "created_at": "2026-07-14T00:50:00Z",
    "credential_id": "",
    "credential_kind": "digitalocean_oauth",
    "granted_at": "2026-07-14T00:52:10Z",
    "id": "39f9f768-a7f9-4c91-9557-3328c53dd054",
    "network": null,
    "oauth": {
      "granted_at": "2026-07-14T00:52:10Z",
      "scopes": [
        "write:jira-work"
      ]
    },
    "provider": "jira",
    "provider_display_name": "Jira",
    "revoked_at": "2026-08-01T09:00:00Z",
    "scopes": [
      "write:jira-work"
    ],
    "status": "revoked",
    "updated_at": "2026-08-01T09:00:00Z",
    "user_id": "alice"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List MCP Servers

`/v2/action-gateway/mcp-servers`

Returns every MCP server your team has registered, sorted by `serverRef`.

##### Request: `/v2/action-gateway/mcp-servers`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`mcpServers` array of object optional

Servers sorted by `serverRef`.

**Show child properties**

`createdAt` string optional

Example: `2026-09-20T15:00:00Z`

When the server was registered, in RFC 3339 format.

`credentialRef` string optional

Example: `action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c`

The secret reference supplied at registration when source=secret, or the team's OAuth credential ID when source=connection. Empty when the server was registered with an `api_key`, whose storage DigitalOcean manages.

`credentialRefSource` string, one of: none, secret, connection optional

Example: `secret`

How requests to the server authenticate: none, secret (a key or token sent in the Authorization header), or connection (each user's own OAuth authorization).

`description` string optional

Example: `Internal documentation search.`

Team-authored description, shown on the server's catalog card. A resync does not replace it.

`endpoint` string optional

Example: `https://mcp.example.com/mcp`

HTTPS URL of the server's MCP endpoint.

`lastSyncedAt` string optional

Example: `2026-09-20T15:04:05Z`

When discovery last succeeded, in RFC 3339 format; empty until the first success.

`oauth_authorization_ttl_seconds` string (uint64) optional

Example: `0`

How long a user's authorization is reused before re-consent. Echoed so a team can read the window without re-registering to discover it.

`oauth_authorize_url` string optional

OAuth authorization endpoint. Set only when `credentialRefSource` is connection; the client secret is never returned.

`oauth_scopes` array of string optional

Example: `[]`

OAuth scopes requested from each user. Set only when `credentialRefSource` is connection.

`protocolVersion` string optional

Example: `2025-06-18`

MCP protocol revision negotiated with the server.

`serverRef` string optional

Example: `docs`

Server identifier, unique within your team. The server's tool slugs are `<server_ref>_<name>`.

`syncError` string optional

Why the latest discovery failed; empty after a successful one.

`syncStatus` string, one of: pending, ok, failed, unsupported_protocol optional

Example: `ok`

Outcome of the latest discovery: pending (no discovery has finished yet), ok, failed, or `unsupported_protocol`. It is not reset while a resync runs.

`toolCount` integer (int32) optional

Example: `3`

Number of tools discovered on the server, whether enabled or not.

`transport` string, one of: streamable_http optional

Example: `streamable_http`

Always `streamable_http`.

`updatedAt` string optional

Example: `2026-09-20T15:04:05Z`

When the server was last modified, in RFC 3339 format.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "mcpServers": [
    {
      "createdAt": "2026-09-20T15:00:00Z",
      "credentialRef": "action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c",
      "credentialRefSource": "secret",
      "description": "Internal documentation search.",
      "endpoint": "https://mcp.example.com/mcp",
      "lastSyncedAt": "2026-09-20T15:04:05Z",
      "oauth_authorization_ttl_seconds": "0",
      "oauth_authorize_url": "",
      "oauth_scopes": [],
      "protocolVersion": "2025-06-18",
      "serverRef": "docs",
      "syncError": "",
      "syncStatus": "ok",
      "toolCount": 2,
      "transport": "streamable_http",
      "updatedAt": "2026-09-20T15:04:05Z"
    }
  ]
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Register an MCP Server

`/v2/action-gateway/mcp-servers`

Registers a remote MCP server for your team, using the streamable HTTP transport. The server starts with `syncStatus` pending and no tools: discover its tools with `POST .../mcp-servers/{server_ref}/resync`, then choose which to enable with `PUT .../mcp-servers/{server_ref}/tools`. The endpoint must be HTTPS and resolve only to public IP addresses. The number of servers per team is limited (10 by default); registering beyond it returns 429. Returns 409 when `serverRef` is already in use.

#### Request Body: `application/json`

`api_key` string optional write-only

Example: `sk_live_example`

For `credentialRefSource` secret: the key or token itself. DigitalOcean stores it; it is write-only and no response returns it. Set exactly one of `credentialRef` and `api_key`.

`credentialRef` string optional

For `credentialRefSource` secret: a reference to a secret your team stores with DigitalOcean. Must match `^`\[A-Za-z0-9]\`\`\[A-Za-z0-9:/_.-]`{0,254}$`. Set exactly one of `credentialRef` and `api_key`.

`credentialRefSource` string, one of: none, secret, connection optional

Example: `secret`

How requests to the server authenticate: none (the default), secret, or connection. With secret, the value from `credentialRef` or `api_key` is sent in the Authorization header, prefixed with "Bearer " unless it already contains a space.

`description` string optional

Example: `Internal documentation search.`

Optional description shown on the server's catalog card. At most 1024 characters.

`endpoint` string required

Example: `https://mcp.example.com/mcp`

HTTPS URL of the server's MCP endpoint. Its host must resolve only to public IP addresses. Required.

`oauth_authorization_ttl_seconds` string (uint64) optional

Example: `86400`

How long, in seconds, a user's authorization may be reused before they must consent again. Optional; 0 means 30 days, and the maximum is 100 years.

Authorizations are not refreshed automatically, so this window, not the provider's token lifetime, decides how often a user is sent back through consent. Set it no longer than the provider's own token lifetime, so the prompt arrives before a tool call fails against an expired token.

`oauth_authorize_url` string optional

Required for `credentialRefSource` connection: the OAuth authorization endpoint. It must use the endpoint's scheme and port, on the endpoint's host or another host under the same registrable domain.

`oauth_client_id` string optional

Required for `credentialRefSource` connection: the client ID of your team's OAuth client for the server. Each user of the server then authorizes individually.

`oauth_client_secret` string optional write-only

Required for `credentialRefSource` connection. Write-only; no response returns it.

`oauth_scopes` array of string optional

Example: `[]`

Required for `credentialRefSource` connection: the scopes to request from each user, 1 to 64 of them.

`oauth_token_url` string optional

Required for `credentialRefSource` connection: the OAuth token endpoint, under the same rules as `oauth_authorize_url`.

`serverRef` string required

Example: `docs`

Identifier for the server, unique within your team. Must match `^`\[a-z]\`\`\[a-z0-9-]`{0,63}$`.

`transport` string, one of: streamable_http optional

Example: `streamable_http`

Optional. `streamable_http`, the default and only accepted value.

##### Request: `/v2/action-gateway/mcp-servers`

### Payload

Content type `application/json`

```json
{
  "api_key": "sk_live_example",
  "credentialRefSource": "secret",
  "description": "Internal documentation search.",
  "endpoint": "https://mcp.example.com/mcp",
  "serverRef": "docs"
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"serverRef":"docs","endpoint":"https://mcp.example.com/mcp","credentialRefSource":"secret","api_key":"sk_live_example","description":"Internal documentation search."}' \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers"
```

#### Responses

**200** The MCP server was registered.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`mcpServer` object optional

The MCP server.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**409** The request could not be completed due to a conflict.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "mcpServer": {
    "createdAt": "2026-09-20T15:00:00Z",
    "credentialRef": "action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c",
    "credentialRefSource": "secret",
    "description": "Internal documentation search.",
    "endpoint": "https://mcp.example.com/mcp",
    "lastSyncedAt": "",
    "oauth_authorization_ttl_seconds": "0",
    "oauth_authorize_url": "",
    "oauth_scopes": [],
    "protocolVersion": "",
    "serverRef": "docs",
    "syncError": "",
    "syncStatus": "pending",
    "toolCount": 0,
    "transport": "streamable_http",
    "updatedAt": "2026-09-20T15:00:00Z"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**409**

```json
{
  "id": "conflict",
  "message": "The request could not be completed due to a conflict."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Get an MCP Server

`/v2/action-gateway/mcp-servers/{server_ref}`

Returns one of your team's MCP servers, or 404 when your team has none with this `serverRef`.

#### Path Parameters

`server_ref` string required

Example: `docs`

Server identifier (`serverRef`) of one of your team's MCP servers.

##### Request: `/v2/action-gateway/mcp-servers/{server_ref}`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers/docs"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`mcpServer` object optional

The MCP server.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "mcpServer": {
    "createdAt": "2026-09-20T15:00:00Z",
    "credentialRef": "action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c",
    "credentialRefSource": "secret",
    "description": "Internal documentation search.",
    "endpoint": "https://mcp.example.com/mcp",
    "lastSyncedAt": "2026-09-20T15:04:05Z",
    "oauth_authorization_ttl_seconds": "0",
    "oauth_authorize_url": "",
    "oauth_scopes": [],
    "protocolVersion": "2025-06-18",
    "serverRef": "docs",
    "syncError": "",
    "syncStatus": "ok",
    "toolCount": 2,
    "transport": "streamable_http",
    "updatedAt": "2026-09-20T15:04:05Z"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## PATCH Update an MCP Server

`/v2/action-gateway/mcp-servers/{server_ref}`

Changes the server's description, which is shown on its catalog card immediately. The endpoint and credentials cannot be changed; to change them, delete the server and register it again.

#### Path Parameters

`server_ref` string required

Example: `docs`

Server identifier (`serverRef`) of one of your team's MCP servers.

#### Request Body: `application/json`

`description` string optional

Example: `Internal documentation and runbook search.`

Replaces the stored description; empty clears it. At most 1024 characters.

##### Request: `/v2/action-gateway/mcp-servers/{server_ref}`

### Payload

Content type `application/json`

```json
{
  "description": "Internal documentation and runbook search."
}
```

### cURL

```bash
curl -X PATCH \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"description":"Internal documentation and runbook search."}' \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers/docs"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`mcpServer` object optional

The MCP server.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "mcpServer": {
    "createdAt": "2026-09-20T15:00:00Z",
    "credentialRef": "action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c",
    "credentialRefSource": "secret",
    "description": "Internal documentation and runbook search.",
    "endpoint": "https://mcp.example.com/mcp",
    "lastSyncedAt": "2026-09-20T15:04:05Z",
    "oauth_authorization_ttl_seconds": "0",
    "oauth_authorize_url": "",
    "oauth_scopes": [],
    "protocolVersion": "2025-06-18",
    "serverRef": "docs",
    "syncError": "",
    "syncStatus": "ok",
    "toolCount": 2,
    "transport": "streamable_http",
    "updatedAt": "2026-09-21T10:00:00Z"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## DELETE Delete an MCP Server

`/v2/action-gateway/mcp-servers/{server_ref}`

Disables all of the server's tools, removes them from your team's catalog, and deletes the server. An `api_key` supplied at registration is deleted; a `credentialRef` you supplied is left untouched. For a server registered with `credentialRefSource` connection, every user connection made through its OAuth client is revoked. Returns the deleted server, or 404 when it does not exist.

#### Path Parameters

`server_ref` string required

Example: `docs`

Server identifier (`serverRef`) of one of your team's MCP servers.

##### Request: `/v2/action-gateway/mcp-servers/{server_ref}`

### cURL

```bash
curl -X DELETE \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers/docs"
```

#### Responses

**200** The MCP server was deleted.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`mcpServer` object optional

The MCP server.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "mcpServer": {
    "createdAt": "2026-09-20T15:00:00Z",
    "credentialRef": "action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c",
    "credentialRefSource": "secret",
    "description": "Internal documentation search.",
    "endpoint": "https://mcp.example.com/mcp",
    "lastSyncedAt": "2026-09-20T15:04:05Z",
    "oauth_authorization_ttl_seconds": "0",
    "oauth_authorize_url": "",
    "oauth_scopes": [],
    "protocolVersion": "2025-06-18",
    "serverRef": "docs",
    "syncError": "",
    "syncStatus": "ok",
    "toolCount": 2,
    "transport": "streamable_http",
    "updatedAt": "2026-09-20T15:04:05Z"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Discover an MCP Server's Tools

`/v2/action-gateway/mcp-servers/{server_ref}/resync`

Connects to the server, lists its tools, and records them. Discovered tools are not callable until you enable them.

When discovery finishes in time, returns 200 with the server and its tools; a discovery failure is reported in `syncStatus` and `syncError` rather than as an error. Otherwise returns 202 with pending set while discovery continues: poll `GET .../mcp-servers/{server_ref}` until `lastSyncedAt` or `syncError` changes (`syncStatus` is not reset while discovery runs).

For a server registered with `credentialRefSource` connection, `user_id` is required; if that user has not authorized the server yet, the response carries authorization with the URL to send them to. Resyncs are rate limited per team, and a team can run only a few discoveries at once; either limit returns 429.

#### Path Parameters

`server_ref` string required

Example: `docs`

Server identifier (`serverRef`) of one of your team's MCP servers.

#### Request Body: `application/json`

`user_id` string optional

Example: `alice`

Required for a server registered with `credentialRefSource` connection: discovery reaches the server as this user, through their connection, because such a server has no team-wide credential. Ignored otherwise.

##### Request: `/v2/action-gateway/mcp-servers/{server_ref}/resync`

### Payload

Content type `application/json`

```json
{}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{}' \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers/docs/resync"
```

#### Responses

**200** Discovery finished.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`authorization` object optional

Set when discovery could not run because `user_id` has no authorized connection to this server yet. It carries the remedy — send the user to `connect_url` — so the caller acts on this response rather than parsing the server's `syncError`.

`mcpServer` object optional

The server, including `syncStatus` and `syncError`.

`pending` boolean optional

Example: `false`

True when discovery is still running (HTTP 202). Poll the server to see the outcome.

`tools` array of object optional

Every tool discovered on the server when discovery finished in time; empty when pending is true.

**Show child properties**

`description` string optional

Example: `Search the documentation index.`

Tool description as the server reports it.

`enabled` boolean optional

Example: `true`

Whether the tool is enabled in your team's catalog.

`name` string optional

Example: `search`

Tool name as the server reports it, normalized to the catalog's naming rules.

`quarantineReason` string optional

Why the tool was quarantined; empty otherwise.

`quarantined` boolean optional

Example: `false`

True when the enabled tool was suspended because it disappeared from the server or its input schema changed incompatibly. Calls to it fail until it is enabled again.

`toolSlug` string optional

Example: `docs_search`

`<server_ref>_<name>`: the tool's catalog slug, and the value to list in `enabledToolSlugs`.

**202** Discovery is still running. Poll the MCP server for its syncStatus.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`authorization` object optional

Set when discovery could not run because `user_id` has no authorized connection to this server yet. It carries the remedy — send the user to `connect_url` — so the caller acts on this response rather than parsing the server's `syncError`.

`mcpServer` object optional

The server, including `syncStatus` and `syncError`.

`pending` boolean optional

Example: `false`

True when discovery is still running (HTTP 202). Poll the server to see the outcome.

`tools` array of object optional

Every tool discovered on the server when discovery finished in time; empty when pending is true.

**Show child properties**

`description` string optional

Example: `Search the documentation index.`

Tool description as the server reports it.

`enabled` boolean optional

Example: `true`

Whether the tool is enabled in your team's catalog.

`name` string optional

Example: `search`

Tool name as the server reports it, normalized to the catalog's naming rules.

`quarantineReason` string optional

Why the tool was quarantined; empty otherwise.

`quarantined` boolean optional

Example: `false`

True when the enabled tool was suspended because it disappeared from the server or its input schema changed incompatibly. Calls to it fail until it is enabled again.

`toolSlug` string optional

Example: `docs_search`

`<server_ref>_<name>`: the tool's catalog slug, and the value to list in `enabledToolSlugs`.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "authorization": null,
  "mcpServer": {
    "createdAt": "2026-09-20T15:00:00Z",
    "credentialRef": "action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c",
    "credentialRefSource": "secret",
    "description": "Internal documentation search.",
    "endpoint": "https://mcp.example.com/mcp",
    "lastSyncedAt": "2026-09-20T15:04:05Z",
    "oauth_authorization_ttl_seconds": "0",
    "oauth_authorize_url": "",
    "oauth_scopes": [],
    "protocolVersion": "2025-06-18",
    "serverRef": "docs",
    "syncError": "",
    "syncStatus": "ok",
    "toolCount": 2,
    "transport": "streamable_http",
    "updatedAt": "2026-09-20T15:04:05Z"
  },
  "pending": false,
  "tools": [
    {
      "description": "Fetch one documentation page.",
      "enabled": false,
      "name": "get_page",
      "quarantineReason": "",
      "quarantined": false,
      "toolSlug": "docs_get_page"
    },
    {
      "description": "Search the documentation index.",
      "enabled": false,
      "name": "search",
      "quarantineReason": "",
      "quarantined": false,
      "toolSlug": "docs_search"
    }
  ]
}
```

**202**

```json
{
  "authorization": null,
  "mcpServer": {
    "createdAt": "2026-09-20T15:00:00Z",
    "credentialRef": "action-infra/tool-credentials/mcp/docs/3f5b2c1e-9a7d-4e8b-b6c4-2d1f0e9a8b7c",
    "credentialRefSource": "secret",
    "description": "Internal documentation search.",
    "endpoint": "https://mcp.example.com/mcp",
    "lastSyncedAt": "2026-09-20T15:04:05Z",
    "oauth_authorization_ttl_seconds": "0",
    "oauth_authorize_url": "",
    "oauth_scopes": [],
    "protocolVersion": "2025-06-18",
    "serverRef": "docs",
    "syncError": "",
    "syncStatus": "pending",
    "toolCount": 2,
    "transport": "streamable_http",
    "updatedAt": "2026-09-20T15:04:05Z"
  },
  "pending": true,
  "tools": []
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List an MCP Server's Tools

`/v2/action-gateway/mcp-servers/{server_ref}/tools`

Returns every tool discovered on the server, sorted by name, with whether each is enabled. Tools that have since disappeared from the server stay listed.

#### Path Parameters

`server_ref` string required

Example: `docs`

Server identifier (`serverRef`) of one of your team's MCP servers.

##### Request: `/v2/action-gateway/mcp-servers/{server_ref}/tools`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers/docs/tools"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`tools` array of object optional

Tools sorted by name.

**Show child properties**

`description` string optional

Example: `Search the documentation index.`

Tool description as the server reports it.

`enabled` boolean optional

Example: `true`

Whether the tool is enabled in your team's catalog.

`name` string optional

Example: `search`

Tool name as the server reports it, normalized to the catalog's naming rules.

`quarantineReason` string optional

Why the tool was quarantined; empty otherwise.

`quarantined` boolean optional

Example: `false`

True when the enabled tool was suspended because it disappeared from the server or its input schema changed incompatibly. Calls to it fail until it is enabled again.

`toolSlug` string optional

Example: `docs_search`

`<server_ref>_<name>`: the tool's catalog slug, and the value to list in `enabledToolSlugs`.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "tools": [
    {
      "description": "Fetch one documentation page.",
      "enabled": true,
      "name": "get_page",
      "quarantineReason": "",
      "quarantined": false,
      "toolSlug": "docs_get_page"
    },
    {
      "description": "Search the documentation index.",
      "enabled": true,
      "name": "search",
      "quarantineReason": "",
      "quarantined": false,
      "toolSlug": "docs_search"
    }
  ]
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## PUT Set Which of an MCP Server's Tools Are Enabled

`/v2/action-gateway/mcp-servers/{server_ref}/tools`

`enabledToolSlugs` is the complete set to enable; every other tool of the server is disabled. An enabled tool is added to your team's catalog at its latest discovered version, and a disabled one is removed from it. Repeating the same set changes nothing.

Before changing anything, each tool being enabled, re-enabled, or moved to a newer discovered version is verified against the live server: it must still be listed there, with the same input schema as when it was discovered. If any check fails or the server cannot be reached, the request fails with 400 and nothing changes; resync and try again. An unknown tool slug is also rejected with 400. A request that needs verification counts against the resync rate limit (429). Returns the server's full tool list.

#### Path Parameters

`server_ref` string required

Example: `docs`

Server identifier (`serverRef`) of one of your team's MCP servers.

#### Request Body: `application/json`

`enabledToolSlugs` array of string required

Example: `["docs_search","docs_get_page"]`

The complete set of tool slugs (`<server_ref>_<name>`) to enable; every other tool of the server is disabled. Empty disables every tool.

`user_id` string optional

Required for a server registered with `credentialRefSource` connection when a tool needs verification against the live server, which can only be reached as a user who has authorized it. Ignored otherwise.

##### Request: `/v2/action-gateway/mcp-servers/{server_ref}/tools`

### Payload

Content type `application/json`

```json
{
  "enabledToolSlugs": [
    "docs_get_page"
  ]
}
```

### cURL

```bash
curl -X PUT \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"enabledToolSlugs":["docs_get_page"]}' \
  "https://api.digitalocean.com/v2/action-gateway/mcp-servers/docs/tools"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`tools` array of object optional

Tools sorted by name.

**Show child properties**

`description` string optional

Example: `Search the documentation index.`

Tool description as the server reports it.

`enabled` boolean optional

Example: `true`

Whether the tool is enabled in your team's catalog.

`name` string optional

Example: `search`

Tool name as the server reports it, normalized to the catalog's naming rules.

`quarantineReason` string optional

Why the tool was quarantined; empty otherwise.

`quarantined` boolean optional

Example: `false`

True when the enabled tool was suspended because it disappeared from the server or its input schema changed incompatibly. Calls to it fail until it is enabled again.

`toolSlug` string optional

Example: `docs_search`

`<server_ref>_<name>`: the tool's catalog slug, and the value to list in `enabledToolSlugs`.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "tools": [
    {
      "description": "Fetch one documentation page.",
      "enabled": true,
      "name": "get_page",
      "quarantineReason": "",
      "quarantined": false,
      "toolSlug": "docs_get_page"
    },
    {
      "description": "Search the documentation index.",
      "enabled": false,
      "name": "search",
      "quarantineReason": "",
      "quarantined": false,
      "toolSlug": "docs_search"
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Output Views

`/v2/action-gateway/output-views`

Returns the output views your team can use: views DigitalOcean publishes to every team, plus your team's own. Cursor-paged with `page_size` and `page_token`; no total is reported.

#### Query Parameters

`tool` string optional

Example: `exa_web_search`

At most one of tool or `tool_id`, as on create. With one set, the list is the published views plus your team's own on that tool version. With neither, it is every live view your team owns plus every published view on a tool version your catalog exposes, across tools.

`tool_id` string optional

Example: `tul_exa_web_search`

Opaque ID of the tool version; see tool.

`page_size` integer 1 – 100 optional

Example: `20`

Page size, 1 to 100. Defaults to 20 when omitted or 0; other values are rejected with 400.

Default: `20`

`page_token` string optional

Example: `ZXhhX3dlYl9zZWFyY2hAMQ`

`next_page_token` from the previous response. Valid only with the same tool selection; otherwise the request is rejected with 400.

##### Request: `/v2/action-gateway/output-views`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/output-views?tool=exa_web_search"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`next_page_token` string optional

Empty on the last page.

`tool` string optional

Example: `exa_web_search`

Tool slug of that version. Empty when the request named no tool.

`tool_id` string optional

Example: `tul_exa_web_search`

The tool version the views are bound to, echoed so an unpinned request learns which version it resolved. Empty when the request named no tool.

`version` string optional

Example: `v1`

Version of that tool, for example `v3`. Empty when the request named no tool.

`views` array of object optional

Ordered by tool version, then published views before your team's own, then name.

**Show child properties**

`audit` object optional Nullable

Null on a preview, which stores nothing.

**Show child properties**

`createdAt` string (date-time) optional Nullable

Example: `2026-09-01T16:00:00Z`

When the resource was created.

`deletedAt` string (date-time) optional Nullable

Set once the resource is deleted.

`updatedAt` string (date-time) optional Nullable

Example: `2026-09-01T16:00:00Z`

When the resource was last modified.

`description` string optional

Example: `Keep only result titles and URLs.`

View description.

`fields` array of string optional

Example: `["results.title","results.url"]`

The dotted output paths a projection keeps; arrays are traversed element-wise. Empty for a transform.

`kind` string, one of: OUTPUT_VIEW_KIND_PROJECTION, OUTPUT_VIEW_KIND_TRANSFORM optional

Example: `OUTPUT_VIEW_KIND_PROJECTION`

`OUTPUT_VIEW_KIND_PROJECTION` or `OUTPUT_VIEW_KIND_TRANSFORM`.

`name` string optional

Example: `titles-and-urls`

View name, unique per tool version among its owner's views.

`output_schema` object optional Nullable

The JSON Schema every result of this view satisfies.

`team_id` string optional

Example: `123456`

Your team's ID on your team's views, and empty for a view DigitalOcean publishes to every team.

`tool` string optional

Example: `exa_web_search`

The provider-qualified tool slug, for example `exa_search`.

`tool_id` string optional

Example: `tul_exa_web_search`

The opaque identity of the exact tool version the view is bound to; tool and version are its readable coordinates.

`version` string optional

Example: `v1`

The tool version, for example `v3`.

`view_id` string optional

Example: `ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7`

Output view ID, for example `ov_` followed by 32 hex digits. Empty on a preview.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "next_page_token": "",
  "tool": "exa_web_search",
  "tool_id": "tul_exa_web_search",
  "version": "v1",
  "views": [
    {
      "audit": {
        "createdAt": "2026-09-01T16:00:00Z",
        "deletedAt": null,
        "updatedAt": "2026-09-01T16:00:00Z"
      },
      "description": "Keep only result titles and URLs.",
      "fields": [
        "results.title",
        "results.url"
      ],
      "kind": "OUTPUT_VIEW_KIND_PROJECTION",
      "name": "titles-and-urls",
      "output_schema": {
        "properties": {
          "results": {
            "items": {
              "properties": {
                "title": {
                  "type": "string"
                },
                "url": {
                  "type": "string"
                }
              },
              "type": "object"
            },
            "type": "array"
          }
        },
        "type": "object"
      },
      "team_id": "123456",
      "tool": "exa_web_search",
      "tool_id": "tul_exa_web_search",
      "version": "v1",
      "view_id": "ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7"
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Create an Output View

`/v2/action-gateway/output-views`

An output view is a named projection of one tool version's output: tool calls that use it return only the listed fields, and its `output_schema` describes the result. The view is bound to the tool version your catalog currently releases. Returns 409 when your team already has a view with this name on the tool version, and 429 when your team already has the maximum of 20 views on it.

#### Request Body: `application/json`

`description` string optional

Example: `Keep only result titles and URLs.`

Optional description. At most 512 bytes.

`fields` array of string required

Example: `["results.title","results.url"]`

Output paths to keep, 1 to 64 of them. Each is a dotted path of up to 8 segments made of `[A-Za-z0-9_-]`, and must be declared by the tool's output schema; arrays are traversed element-wise. Paths may not repeat or be a prefix of one another. The tool version must declare an object output schema.

`name` string required

Example: `titles-and-urls`

View name. Must match `^`\[a-z]\`\`\[a-z0-9_-]`{0,63}$` and be unique per tool version within your team.

`tool` string optional

Example: `exa_web_search`

Exactly one of tool or `tool_id` selects the tool version. tool is a provider-qualified slug, optionally pinned as `<tool>@<version>`. You can bind only the released version your catalog exposes: an unpinned slug resolves it and the response echoes it as version; a pin must equal it, and a `tool_id` (the opaque identity from tool search) must be that version. Any other version returns 404.

`tool_id` string optional

Opaque ID of the tool version; see tool.

##### Request: `/v2/action-gateway/output-views`

### Payload

Content type `application/json`

```json
{
  "description": "Keep only result titles and URLs.",
  "fields": [
    "results.title",
    "results.url"
  ],
  "name": "titles-and-urls",
  "tool": "exa_web_search"
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"tool":"exa_web_search","name":"titles-and-urls","description":"Keep only result titles and URLs.","fields":["results.title","results.url"]}' \
  "https://api.digitalocean.com/v2/action-gateway/output-views"
```

#### Responses

**200** The output view was created.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`view` object optional

The output view.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**409** The request could not be completed due to a conflict.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "view": {
    "audit": {
      "createdAt": "2026-09-01T16:00:00Z",
      "deletedAt": null,
      "updatedAt": "2026-09-01T16:00:00Z"
    },
    "description": "Keep only result titles and URLs.",
    "fields": [
      "results.title",
      "results.url"
    ],
    "kind": "OUTPUT_VIEW_KIND_PROJECTION",
    "name": "titles-and-urls",
    "output_schema": {
      "properties": {
        "results": {
          "items": {
            "properties": {
              "title": {
                "type": "string"
              },
              "url": {
                "type": "string"
              }
            },
            "type": "object"
          },
          "type": "array"
        }
      },
      "type": "object"
    },
    "team_id": "123456",
    "tool": "exa_web_search",
    "tool_id": "tul_exa_web_search",
    "version": "v1",
    "view_id": "ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**409**

```json
{
  "id": "conflict",
  "message": "The request could not be completed due to a conflict."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Preview an Output View

`/v2/action-gateway/output-views/preview`

Validates a projection exactly as create does and returns the view it would produce, including `output_schema`, without storing anything. name and description are optional here; name uniqueness and the per-tool-version limit are not checked.

#### Request Body: `application/json`

`description` string optional

Example: `Keep only result titles and URLs.`

Optional description, validated as on create when present.

`fields` array of string required

Example: `["results.title","results.url"]`

Output paths to keep; as on create.

`name` string optional

Example: `titles-and-urls`

Optional view name, validated as on create when present.

`tool` string optional

Example: `exa_web_search`

Tool slug, optionally pinned; as on create.

`tool_id` string optional

Opaque ID of the tool version; as on create.

##### Request: `/v2/action-gateway/output-views/preview`

### Payload

Content type `application/json`

```json
{
  "fields": [
    "results.title",
    "results.url"
  ],
  "tool": "exa_web_search"
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"tool":"exa_web_search","fields":["results.title","results.url"]}' \
  "https://api.digitalocean.com/v2/action-gateway/output-views/preview"
```

#### Responses

**200** The output view is valid.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`view` object optional

The output view.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "view": {
    "audit": null,
    "description": "Keep only result titles and URLs.",
    "fields": [
      "results.title",
      "results.url"
    ],
    "kind": "OUTPUT_VIEW_KIND_PROJECTION",
    "name": "titles-and-urls",
    "output_schema": {
      "properties": {
        "results": {
          "items": {
            "properties": {
              "title": {
                "type": "string"
              },
              "url": {
                "type": "string"
              }
            },
            "type": "object"
          },
          "type": "array"
        }
      },
      "type": "object"
    },
    "team_id": "123456",
    "tool": "exa_web_search",
    "tool_id": "tul_exa_web_search",
    "version": "v1",
    "view_id": ""
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Get an Output View

`/v2/action-gateway/output-views/{view_id}`

Returns one of your team's views, or a published view on a tool version your catalog exposes. Returns 404 when the view does not exist, was deleted, or belongs to another team.

#### Path Parameters

`view_id` string required

Example: `ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7`

Output view ID.

##### Request: `/v2/action-gateway/output-views/{view_id}`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/output-views/ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`view` object optional

The output view.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "view": {
    "audit": {
      "createdAt": "2026-09-01T16:00:00Z",
      "deletedAt": null,
      "updatedAt": "2026-09-01T16:00:00Z"
    },
    "description": "Keep only result titles and URLs.",
    "fields": [
      "results.title",
      "results.url"
    ],
    "kind": "OUTPUT_VIEW_KIND_PROJECTION",
    "name": "titles-and-urls",
    "output_schema": {
      "properties": {
        "results": {
          "items": {
            "properties": {
              "title": {
                "type": "string"
              },
              "url": {
                "type": "string"
              }
            },
            "type": "object"
          },
          "type": "array"
        }
      },
      "type": "object"
    },
    "team_id": "123456",
    "tool": "exa_web_search",
    "tool_id": "tul_exa_web_search",
    "version": "v1",
    "view_id": "ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## DELETE Delete an Output View

`/v2/action-gateway/output-views/{view_id}`

Deletes one of your team's views and returns it with audit.deleted_at set; it can no longer be used. Published views cannot be deleted. Returns 404 when the view does not exist, is already deleted, or is not your team's.

#### Path Parameters

`view_id` string required

Example: `ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7`

Output view ID.

##### Request: `/v2/action-gateway/output-views/{view_id}`

### cURL

```bash
curl -X DELETE \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/output-views/ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7"
```

#### Responses

**200** The output view was deleted.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`view` object optional

The output view.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "view": {
    "audit": {
      "createdAt": "2026-09-01T16:00:00Z",
      "deletedAt": "2026-09-10T08:00:00Z",
      "updatedAt": "2026-09-10T08:00:00Z"
    },
    "description": "Keep only result titles and URLs.",
    "fields": [
      "results.title",
      "results.url"
    ],
    "kind": "OUTPUT_VIEW_KIND_PROJECTION",
    "name": "titles-and-urls",
    "output_schema": {
      "properties": {
        "results": {
          "items": {
            "properties": {
              "title": {
                "type": "string"
              },
              "url": {
                "type": "string"
              }
            },
            "type": "object"
          },
          "type": "array"
        }
      },
      "type": "object"
    },
    "team_id": "123456",
    "tool": "exa_web_search",
    "tool_id": "tul_exa_web_search",
    "version": "v1",
    "view_id": "ov_4f0c2a9d8e1b47c6a3d5f7e9b1c3a5d7"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Sessions

`/v2/action-gateway/sessions`

Returns the sessions you can see, newest first. Offset-paged with page and `per_page`; `end_user_id` filters to one actor.

#### Query Parameters

`end_user_id` string optional

Example: `alice`

Optional `actor_id`, matched exactly.

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

##### Request: `/v2/action-gateway/sessions`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/sessions"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`pagination` object optional

Paging applied to this response and the total number of sessions.

`sessions` array of object optional

Sessions, newest first.

**Show child properties**

`actorId` string optional

Example: `alice`

Empty when the session is not bound to an actor.

`agentName` string optional

Name of the agent that started the session. Empty for sessions created through this API.

`agentUrn` string optional

URN of the agent that started the session. Empty for sessions created through this API.

`config` object optional

Example: `{"outputViews":{"exa_web_search":"titles-and-urls"},"preloadTools":["exa_web_search"]}`

Session options as supplied at creation. Omitted when none were supplied.

`createdAt` string (date-time) optional Nullable

Example: `2026-09-24T12:00:00Z`

When the session was created.

`insights` object optional

Omitted when no explicit customer Insights choice was stored.

`name` string optional

Example: `support-bot-alice`

The required human-readable session name.

`network` object optional

Omitted when the request omitted network. A present binding currently supports only VPC attachment by UUID.

`owning_user_id` string optional

Example: `2f8e1c3a-6b4d-4f7e-9a1c-5d3b7e9f1a2c`

DigitalOcean user ID of the user who created the session, when recorded.

`policy` object optional

The session's tool-permission policy.

`sessionUrn` string optional

Example: `do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b`

Session URN, for example `do:managed_agent_session:<uuid>`.

`tools` object optional

Omitted when the request omitted tools (all tools). A present selection with no references represents tools: \[].

**Show child properties**

`references` array of object optional

The selected references. Empty means the session exposes no tools.

**Show child properties**

`kind` string, one of: SESSION_TOOL_REFERENCE_KIND_TOOL, SESSION_TOOL_REFERENCE_KIND_TOOLBELT optional

Example: `SESSION_TOOL_REFERENCE_KIND_TOOLBELT`

Whether name is a tool or a toolbelt.

`name` string optional

Example: `search-toolbelt`

Tool slug or toolbelt name.

`version` string optional

Example: `1`

Pinned version.

`updatedAt` string (date-time) optional Nullable

Example: `2026-09-24T12:00:00Z`

When the session was last modified.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 1
  },
  "sessions": [
    {
      "actorId": "alice",
      "agentName": "",
      "agentUrn": "",
      "config": {
        "outputViews": {
          "exa_web_search": "titles-and-urls"
        },
        "preloadTools": [
          "exa_web_search"
        ]
      },
      "createdAt": "2026-09-24T12:00:00Z",
      "name": "support-bot-alice",
      "owning_user_id": "2f8e1c3a-6b4d-4f7e-9a1c-5d3b7e9f1a2c",
      "policy": {
        "defaultAction": "ask",
        "rules": [
          {
            "action": "allow",
            "match": {},
            "tool": "toolbelt:search-toolbelt@1"
          },
          {
            "action": "ask",
            "match": {},
            "tool": "jira_create_issue"
          }
        ]
      },
      "sessionUrn": "do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b",
      "tools": {
        "references": [
          {
            "kind": "SESSION_TOOL_REFERENCE_KIND_TOOLBELT",
            "name": "search-toolbelt",
            "version": "1"
          },
          {
            "kind": "SESSION_TOOL_REFERENCE_KIND_TOOL",
            "name": "jira_create_issue",
            "version": "1"
          }
        ]
      },
      "updatedAt": "2026-09-24T12:00:00Z"
    }
  ]
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Create a Session

`/v2/action-gateway/sessions`

Creates a session owned by your team, with a new session URN, and returns the MCP URL an agent connects to. tools selects the tools and toolbelts the session exposes, and policy decides which calls run, need approval, or are refused.

#### Request Body: `application/json`

`actorId` string optional

Example: `alice`

Optional actor binding: the user whose connections the session's tool calls use, matched against connection `user_id`. When set it must use the same alphabet as a connection's `user_id`, `[A-Za-z0-9._-]`, at most 64 characters; anything else is rejected with 400.

`config` object optional

Optional session options. config.preloadTools may contain concrete tool names (optionally version-pinned) and version-pinned toolbelt references. config.customInstructions is a string of at most 1000 characters. config.outputViews maps up to 100 tool slugs to the name of an output view to apply to that tool's results.

`insights` object optional

Omitted when the request omitted insights or explicitly sent null.

`name` string required

Example: `support-bot-alice`

Required human-readable session name.

`network` object optional

Product-level session network binding. When set, `vpc_uuid` must identify a VPC owned by the session's team.

`policy` object optional

Optional tool-permission policy. Omitted asks before every call.

`tools` array of string optional

Example: `["toolbelt:search-toolbelt@1","jira_create_issue"]`

Omitted enables every tool. An explicit empty array enables no tools. Direct tools may be `<tool>` or `<tool>@<version>`, using provider-qualified tool slugs, and a pinned version must be the released one; toolbelt references must be version-pinned as `toolbelt:<belt-name>@<version>`.

##### Request: `/v2/action-gateway/sessions`

### Payload

Content type `application/json`

```json
{
  "actorId": "alice",
  "config": {
    "outputViews": {
      "exa_web_search": "titles-and-urls"
    },
    "preloadTools": [
      "exa_web_search"
    ]
  },
  "name": "support-bot-alice",
  "policy": {
    "defaultAction": "ask",
    "rules": [
      {
        "action": "allow",
        "match": {},
        "tool": "toolbelt:search-toolbelt@1"
      },
      {
        "action": "ask",
        "match": {},
        "tool": "jira_create_issue"
      }
    ]
  },
  "tools": [
    "toolbelt:search-toolbelt@1",
    "jira_create_issue"
  ]
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"name":"support-bot-alice","actorId":"alice","tools":["toolbelt:search-toolbelt@1","jira_create_issue"],"policy":{"defaultAction":"ask","rules":[{"tool":"toolbelt:search-toolbelt@1","match":{},"action":"allow"},{"tool":"jira_create_issue","match":{},"action":"ask"}]},"config":{"preloadTools":["exa_web_search"],"outputViews":{"exa_web_search":"titles-and-urls"}}}' \
  "https://api.digitalocean.com/v2/action-gateway/sessions"
```

#### Responses

**200** The session was created.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`mcpUrl` string optional

Example: `https://actions.do-ai.run/mcp/session/2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b`

URL of the session's MCP endpoint, for the agent to connect to.

`session` object optional

The created session.

`tools` array of string optional

Example: `["toolbelt:search-toolbelt@1","jira_create_issue@1"]`

Canonical, version-pinned selected tool references. Empty when the session exposes every tool.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "mcpUrl": "https://actions.do-ai.run/mcp/session/2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b",
  "session": {
    "actorId": "alice",
    "agentName": "",
    "agentUrn": "",
    "config": {
      "outputViews": {
        "exa_web_search": "titles-and-urls"
      },
      "preloadTools": [
        "exa_web_search"
      ]
    },
    "createdAt": "2026-09-24T12:00:00Z",
    "name": "support-bot-alice",
    "owning_user_id": "2f8e1c3a-6b4d-4f7e-9a1c-5d3b7e9f1a2c",
    "policy": {
      "defaultAction": "ask",
      "rules": [
        {
          "action": "allow",
          "match": {},
          "tool": "toolbelt:search-toolbelt@1"
        },
        {
          "action": "ask",
          "match": {},
          "tool": "jira_create_issue"
        }
      ]
    },
    "sessionUrn": "do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b",
    "tools": {
      "references": [
        {
          "kind": "SESSION_TOOL_REFERENCE_KIND_TOOLBELT",
          "name": "search-toolbelt",
          "version": "1"
        },
        {
          "kind": "SESSION_TOOL_REFERENCE_KIND_TOOL",
          "name": "jira_create_issue",
          "version": "1"
        }
      ]
    },
    "updatedAt": "2026-09-24T12:00:00Z"
  },
  "tools": [
    "toolbelt:search-toolbelt@1",
    "jira_create_issue@1"
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Search Sessions

`/v2/action-gateway/sessions/search`

Returns the sessions you can see whose name, `actor_id`, `session_urn`, or `agentName` contains query, case-insensitively, ranked across the whole result set. Cursor-paged with `page_size` and `page_token`; no total is reported.

#### Query Parameters

`query` string optional

Example: `search`

Matched case-insensitively as a substring of name, `actor_id`, `session_urn`, or `agentName`. Empty matches every session you can see. Results are ordered by match quality (an exact name, `session_urn`, or `actor_id` match, then a prefix, then a substring), then by name, across the whole result set, so a client must not re-sort a page to rank it.

`end_user_id` string optional

Example: `alice`

`end_user_id`, when set, is an exact filter on `actor_id` (the same filter as on the session list) and combines with query as AND.

`page_size` integer 1 – 100 optional

Example: `20`

Page size, 1 to 100. Defaults to 20 when omitted or 0; other values are rejected with 400.

Default: `20`

`page_token` string optional

Example: `ZXhhX3dlYl9zZWFyY2hAMQ`

`next_page_token` from the previous response. Valid only with the same query and `end_user_id`; otherwise the request is rejected with 400.

##### Request: `/v2/action-gateway/sessions/search`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/sessions/search?query=support"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`next_page_token` string optional

Token for the next page; empty on the last page.

`sessions` array of object optional

Matching sessions in rank order.

**Show child properties**

`actorId` string optional

Example: `alice`

Empty when the session is not bound to an actor.

`agentName` string optional

Name of the agent that started the session. Empty for sessions created through this API.

`agentUrn` string optional

URN of the agent that started the session. Empty for sessions created through this API.

`config` object optional

Example: `{"outputViews":{"exa_web_search":"titles-and-urls"},"preloadTools":["exa_web_search"]}`

Session options as supplied at creation. Omitted when none were supplied.

`createdAt` string (date-time) optional Nullable

Example: `2026-09-24T12:00:00Z`

When the session was created.

`insights` object optional

Omitted when no explicit customer Insights choice was stored.

`name` string optional

Example: `support-bot-alice`

The required human-readable session name.

`network` object optional

Omitted when the request omitted network. A present binding currently supports only VPC attachment by UUID.

`owning_user_id` string optional

Example: `2f8e1c3a-6b4d-4f7e-9a1c-5d3b7e9f1a2c`

DigitalOcean user ID of the user who created the session, when recorded.

`policy` object optional

The session's tool-permission policy.

`sessionUrn` string optional

Example: `do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b`

Session URN, for example `do:managed_agent_session:<uuid>`.

`tools` object optional

Omitted when the request omitted tools (all tools). A present selection with no references represents tools: \[].

**Show child properties**

`references` array of object optional

The selected references. Empty means the session exposes no tools.

**Show child properties**

`kind` string, one of: SESSION_TOOL_REFERENCE_KIND_TOOL, SESSION_TOOL_REFERENCE_KIND_TOOLBELT optional

Example: `SESSION_TOOL_REFERENCE_KIND_TOOLBELT`

Whether name is a tool or a toolbelt.

`name` string optional

Example: `search-toolbelt`

Tool slug or toolbelt name.

`version` string optional

Example: `1`

Pinned version.

`updatedAt` string (date-time) optional Nullable

Example: `2026-09-24T12:00:00Z`

When the session was last modified.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "next_page_token": "",
  "sessions": [
    {
      "actorId": "alice",
      "agentName": "",
      "agentUrn": "",
      "config": {
        "outputViews": {
          "exa_web_search": "titles-and-urls"
        },
        "preloadTools": [
          "exa_web_search"
        ]
      },
      "createdAt": "2026-09-24T12:00:00Z",
      "name": "support-bot-alice",
      "owning_user_id": "2f8e1c3a-6b4d-4f7e-9a1c-5d3b7e9f1a2c",
      "policy": {
        "defaultAction": "ask",
        "rules": [
          {
            "action": "allow",
            "match": {},
            "tool": "toolbelt:search-toolbelt@1"
          },
          {
            "action": "ask",
            "match": {},
            "tool": "jira_create_issue"
          }
        ]
      },
      "sessionUrn": "do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b",
      "tools": {
        "references": [
          {
            "kind": "SESSION_TOOL_REFERENCE_KIND_TOOLBELT",
            "name": "search-toolbelt",
            "version": "1"
          },
          {
            "kind": "SESSION_TOOL_REFERENCE_KIND_TOOL",
            "name": "jira_create_issue",
            "version": "1"
          }
        ]
      },
      "updatedAt": "2026-09-24T12:00:00Z"
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## DELETE Delete a Session

`/v2/action-gateway/sessions/{session_urn}`

Deletes the session and its policy. Succeeds without deleting anything when the session does not exist or you cannot delete it, so repeating a delete also succeeds. A malformed `session_urn` is rejected with 400.

#### Path Parameters

`session_urn` string required

Example: `do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b`

URN of the session, for example `do:managed_agent_session:<uuid>`.

##### Request: `/v2/action-gateway/sessions/{session_urn}`

### cURL

```bash
curl -X DELETE \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/sessions/do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b"
```

#### Responses

**200** The session was deleted.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Toolbelts

`/v2/action-gateway/toolbelts`

Returns your team's toolbelts, one entry per name describing its latest version, sorted by name. Offset-paged with page and `per_page`; pagination reports the total.

#### Query Parameters

`status` string, one of: active, deprecated, all optional

Example: `active`

active (default), deprecated, or all. Case-insensitive.

Default: `active`

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

##### Request: `/v2/action-gateway/toolbelts`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`pagination` object optional

Paging applied to this response and the total number of toolbelts.

`toolbelts` array of object optional

Toolbelts sorted by name.

**Show child properties**

`description` string optional

Example: `Tools for searching and fetching public web pages.`

Description of the latest version.

`display_name` string optional

Example: `Search Tools`

Human-readable label of the latest version.

`latest_version` string optional

Example: `1`

Latest version number, as a string.

`name` string optional

Example: `search-toolbelt`

Toolbelt name.

`reference_latest` string optional

Example: `search-toolbelt`

The toolbelt name, which refers to whichever version is latest.

`status` string, one of: active, deprecated optional

Example: `active`

active, or deprecated once the toolbelt is deleted.

`tool_count` integer (int32) optional

Example: `2`

Number of members in the latest version.

`updated_at` string (date-time) optional

Example: `2026-06-11T12:00:00Z`

When the latest version was last modified, in RFC 3339 format.

`version_count` integer (int32) optional

Example: `1`

Number of versions recorded under this name, including versions from before the toolbelt was deleted and the name reused.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 1
  },
  "toolbelts": [
    {
      "description": "Tools for searching and fetching public web pages.",
      "display_name": "Search Tools",
      "latest_version": "1",
      "name": "search-toolbelt",
      "reference_latest": "search-toolbelt",
      "status": "active",
      "tool_count": 2,
      "updated_at": "2026-06-11T12:00:00Z",
      "version_count": 1
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Create a Toolbelt

`/v2/action-gateway/toolbelts`

A toolbelt is a named set of tools, each pinned to a version, that your team can attach to sessions as `toolbelt:<name>@<version>`. The first version is 1 unless version is set. Returns 409 if an active toolbelt with the same name exists. The name of a deleted toolbelt can be reused; its versions start again from the initial version.

#### Request Body: `application/json`

`description` string optional

Example: `Tools for searching and fetching public web pages.`

Optional description. At most 255 bytes.

`display_name` string optional

Example: `Search Tools`

Optional human-readable label, separate from name. At most 128 bytes.

`name` string required

Example: `search-toolbelt`

Toolbelt name, unique among your team's active toolbelts. Must match `^`\[a-z]\`\`\[a-z0-9_-]`{0,63}$`; `search` is reserved.

`tools` array of string optional

Example: `["exa_web_search","exa_web_fetch"]`

Optional initial members, as catalog tool slugs (`<provider>_<name>`). Each may be pinned as `<tool_slug>@<version>`; a pin must equal the tool's current released version, and an unpinned tool is pinned to that version. Every tool must be an active catalog tool. Duplicates are merged; at most 500 tools. Empty creates a toolbelt with no tools.

`version` string optional

Example: `1`

Optional initial version number, a positive integer such as `1`. Defaults to 1.

##### Request: `/v2/action-gateway/toolbelts`

### Payload

Content type `application/json`

```json
{
  "description": "Tools for searching and fetching public web pages.",
  "display_name": "Search Tools",
  "name": "search-toolbelt",
  "tools": [
    "exa_web_search",
    "exa_web_fetch"
  ]
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"name":"search-toolbelt","display_name":"Search Tools","description":"Tools for searching and fetching public web pages.","tools":["exa_web_search","exa_web_fetch"]}' \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts"
```

#### Responses

**200** The toolbelt was created.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`toolbelt` object optional

The first version of the new toolbelt.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**409** The request could not be completed due to a conflict.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "toolbelt": {
    "created_at": "2026-06-11T12:00:00Z",
    "description": "Tools for searching and fetching public web pages.",
    "display_name": "Search Tools",
    "name": "search-toolbelt",
    "reference": "search-toolbelt@1",
    "reference_latest": "search-toolbelt",
    "status": "active",
    "tool_count": 2,
    "tools": [
      "exa_web_fetch@1",
      "exa_web_search@1"
    ],
    "updated_at": "2026-06-11T12:00:00Z",
    "version": "1"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**409**

```json
{
  "id": "conflict",
  "message": "The request could not be completed due to a conflict."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Search Toolbelts

`/v2/action-gateway/toolbelts/search`

Returns your team's toolbelts whose name, display name, or description contains query, case-insensitively, sorted by name. Cursor-paged with `page_size` and `page_token`; no total is reported.

#### Query Parameters

`query` string optional

Example: `search`

Matched case-insensitively as a substring of name, `display_name`, or description. Empty matches every toolbelt of your team.

`status` string, one of: active, deprecated, all optional

Example: `active`

active (default), deprecated, or all. Case-insensitive.

Default: `active`

`page_size` integer 1 – 100 optional

Example: `20`

Page size, 1 to 100. Defaults to 20 when omitted or 0; other values are rejected with 400.

Default: `20`

`page_token` string optional

Example: `ZXhhX3dlYl9zZWFyY2hAMQ`

`next_page_token` from the previous response. Valid only with the same query and status; otherwise the request is rejected with 400.

##### Request: `/v2/action-gateway/toolbelts/search`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts/search?query=search"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`next_page_token` string optional

Token for the next page; empty on the last page.

`toolbelts` array of object optional

Matching toolbelts sorted by name.

**Show child properties**

`description` string optional

Example: `Tools for searching and fetching public web pages.`

Description of the latest version.

`display_name` string optional

Example: `Search Tools`

Human-readable label of the latest version.

`latest_version` string optional

Example: `1`

Latest version number, as a string.

`name` string optional

Example: `search-toolbelt`

Toolbelt name.

`reference_latest` string optional

Example: `search-toolbelt`

The toolbelt name, which refers to whichever version is latest.

`status` string, one of: active, deprecated optional

Example: `active`

active, or deprecated once the toolbelt is deleted.

`tool_count` integer (int32) optional

Example: `2`

Number of members in the latest version.

`updated_at` string (date-time) optional

Example: `2026-06-11T12:00:00Z`

When the latest version was last modified, in RFC 3339 format.

`version_count` integer (int32) optional

Example: `1`

Number of versions recorded under this name, including versions from before the toolbelt was deleted and the name reused.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "next_page_token": "",
  "toolbelts": [
    {
      "description": "Tools for searching and fetching public web pages.",
      "display_name": "Search Tools",
      "latest_version": "1",
      "name": "search-toolbelt",
      "reference_latest": "search-toolbelt",
      "status": "active",
      "tool_count": 2,
      "updated_at": "2026-06-11T12:00:00Z",
      "version_count": 1
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Get a Toolbelt

`/v2/action-gateway/toolbelts/{name}`

Returns the latest version of the named toolbelt, or the version requested, with catalog metadata for one page of its members in `tool_details`. A deleted toolbelt returns 404 unless a version is requested.

#### Path Parameters

`name` string required

Example: `search-toolbelt`

Toolbelt name.

#### Query Parameters

`version` string optional

Example: `1`

Version number to read. Empty returns the latest version.

`page_size` integer 1 – 100 optional

Example: `20`

Page size for `tool_details` only; toolbelt.tools always lists every member. Defaults to 20; values above 100 are capped at 100.

Default: `20`

`page_token` string optional

Example: `ZXhhX3dlYl9zZWFyY2hAMQ`

`next_page_token` from the previous response, to continue `tool_details`. Treat it as opaque, and send it only with the same toolbelt, version, and search.

`search` string optional

Example: `search`

Restricts `tool_details` to members whose tool slug, name, title, description, provider, or category contains this value, case-insensitively. It does not filter toolbelt.tools. This is a substring match, not a filter expression.

##### Request: `/v2/action-gateway/toolbelts/{name}`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts/search-toolbelt"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`next_page_token` string optional

Token for the next page of `tool_details`; empty on the last page.

`tool_details` array of object optional

The requested page of resolved catalog metadata for the members named in toolbelt.tools. It is a page, so it is not the membership of the toolbelt: read toolbelt.tools for that, and toolbelt.tool_count for the total. Ordering matches toolbelt.tools.

**Show child properties**

`category` string optional

Example: `web-search`

Best-effort catalog metadata and is empty for a large share of the catalog. Do not rely on it being present.

`description` string optional

Example: `Search the public web with Exa and return relevant results with normalized text snippets.`

What the tool does.

`name` string optional

Example: `web_search`

The unqualified tool name, without the provider prefix.

`provider` string optional

Example: `exa`

The ID of the provider that offers the tool, broken out so a client never has to split `tool_slug`.

`title` string optional

Example: `Web Search`

Human-readable tool title.

`tool_slug` string optional

Example: `exa_web_search`

The provider-qualified, stable tool identifier (`<provider>_<name>`). Treat it as opaque and pass it back verbatim.

`version` integer (int32) optional

Example: `1`

The version this toolbelt version pins for the member, matching the `@<version>` suffix in the corresponding toolbelt.tools entry.

`toolbelt` object optional

The requested toolbelt version.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "next_page_token": "",
  "tool_details": [
    {
      "category": "web-fetch",
      "description": "Fetch the contents of a public web page with Exa.",
      "name": "web_fetch",
      "provider": "exa",
      "title": "Web Fetch",
      "tool_slug": "exa_web_fetch",
      "version": 1
    },
    {
      "category": "web-search",
      "description": "Search the public web with Exa and return relevant results with normalized text snippets.",
      "name": "web_search",
      "provider": "exa",
      "title": "Web Search",
      "tool_slug": "exa_web_search",
      "version": 1
    }
  ],
  "toolbelt": {
    "created_at": "2026-06-11T12:00:00Z",
    "description": "Tools for searching and fetching public web pages.",
    "display_name": "Search Tools",
    "name": "search-toolbelt",
    "reference": "search-toolbelt@1",
    "reference_latest": "search-toolbelt",
    "status": "active",
    "tool_count": 2,
    "tools": [
      "exa_web_fetch@1",
      "exa_web_search@1"
    ],
    "updated_at": "2026-06-11T12:00:00Z",
    "version": "1"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## DELETE Delete a Toolbelt

`/v2/action-gateway/toolbelts/{name}`

Marks the toolbelt deleted and returns its latest version with status deprecated. Its versions stay readable by explicit version until the name is reused. The name is free to reuse immediately; a new toolbelt with it starts a fresh version history. Returns 404 when the toolbelt does not exist or is already deleted.

#### Path Parameters

`name` string required

Example: `search-toolbelt`

Toolbelt name.

##### Request: `/v2/action-gateway/toolbelts/{name}`

### cURL

```bash
curl -X DELETE \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts/search-toolbelt"
```

#### Responses

**200** The toolbelt was deprecated.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`toolbelt` object optional

After adding or removing tools: the new version, or the current version when membership did not change. After a delete: the latest version, with status deprecated.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "toolbelt": {
    "created_at": "2026-06-11T12:00:00Z",
    "description": "Tools for searching and fetching public web pages.",
    "display_name": "Search Tools",
    "name": "search-toolbelt",
    "reference": "search-toolbelt@1",
    "reference_latest": "search-toolbelt",
    "status": "deprecated",
    "tool_count": 2,
    "tools": [
      "exa_web_fetch@1",
      "exa_web_search@1"
    ],
    "updated_at": "2026-06-15T10:00:00Z",
    "version": "1"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List the Providers in a Toolbelt

`/v2/action-gateway/toolbelts/{name}/providers`

Returns the distinct providers of the toolbelt's members, each with its member count, sorted by provider ID. List one provider's members with `GET .../toolbelts/{name}/providers/{provider}/tools`.

#### Path Parameters

`name` string required

Example: `search-toolbelt`

Toolbelt name.

#### Query Parameters

`version` string optional

Example: `1`

Version number to read. Empty reads the latest version.

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

`search` string optional

Example: `search`

Restricts providers whose ID, name, or description contains this value, case-insensitively.

##### Request: `/v2/action-gateway/toolbelts/{name}/providers`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts/search-toolbelt/providers"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`pagination` object optional

Paging applied to this response and the total number of providers.

`providers` array of object optional

Providers sorted by ID.

**Show child properties**

`categories` array of string optional

Example: `["web-search"]`

The distinct tool categories among this toolbelt's members for the provider (sorted). Empty for `_legacy` or when no member has a category.

`created_at` string (date-time) optional Nullable

Example: `2026-06-01T00:00:00Z`

When the provider was added to the catalog. Unset for `_legacy`.

`description` string optional

Example: `Web search tools backed by Exa.`

Provider description. Empty for `_legacy`.

`id` string optional

Example: `exa`

Equals provider; present so the entry has the same shape as a toolkit.

`name` string optional

Example: `Exa`

The provider's display name. Empty for `_legacy`.

`provider` string optional

Example: `exa`

The provider ID. Members recorded without a provider are grouped under `_legacy`, which is also accepted in the provider path.

`tool_count` integer (int32) optional

Example: `2`

How many toolbelt members belong to this provider.

`toolbelt` object optional

The toolbelt version the providers were read from.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 1
  },
  "providers": [
    {
      "categories": [
        "web-fetch",
        "web-search"
      ],
      "created_at": "2026-06-01T00:00:00Z",
      "description": "Web search tools backed by Exa.",
      "id": "exa",
      "name": "Exa",
      "provider": "exa",
      "tool_count": 2
    }
  ],
  "toolbelt": {
    "created_at": "2026-06-11T12:00:00Z",
    "description": "Tools for searching and fetching public web pages.",
    "display_name": "Search Tools",
    "name": "search-toolbelt",
    "reference": "search-toolbelt@1",
    "reference_latest": "search-toolbelt",
    "status": "active",
    "tool_count": 2,
    "tools": [
      "exa_web_fetch@1",
      "exa_web_search@1"
    ],
    "updated_at": "2026-06-11T12:00:00Z",
    "version": "1"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List a Toolbelt's Tools from One Provider

`/v2/action-gateway/toolbelts/{name}/providers/{provider}/tools`

Returns a page of catalog metadata for the toolbelt's members from the named provider, sorted by tool slug. An unknown provider returns an empty page.

#### Path Parameters

`name` string required

Example: `search-toolbelt`

Toolbelt name.

`provider` string required

Example: `exa`

A provider ID from the toolbelt's provider list, or `_legacy` for members recorded without a provider. Matched exactly.

#### Query Parameters

`version` string optional

Example: `1`

Version number to read. Empty reads the latest version.

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

`search` string optional

Example: `search`

Restricts tools whose slug, name, title, description, provider, or category contains this value, case-insensitively, within the provider only.

##### Request: `/v2/action-gateway/toolbelts/{name}/providers/{provider}/tools`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts/search-toolbelt/providers/exa/tools"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`pagination` object optional

Paging applied to this response and the total number of matching members.

`tools` array of object optional

Members sorted by tool slug.

**Show child properties**

`category` string optional

Example: `web-search`

Best-effort catalog metadata and is empty for a large share of the catalog. Do not rely on it being present.

`description` string optional

Example: `Search the public web with Exa and return relevant results with normalized text snippets.`

What the tool does.

`name` string optional

Example: `web_search`

The unqualified tool name, without the provider prefix.

`provider` string optional

Example: `exa`

The ID of the provider that offers the tool, broken out so a client never has to split `tool_slug`.

`title` string optional

Example: `Web Search`

Human-readable tool title.

`tool_slug` string optional

Example: `exa_web_search`

The provider-qualified, stable tool identifier (`<provider>_<name>`). Treat it as opaque and pass it back verbatim.

`version` integer (int32) optional

Example: `1`

The version this toolbelt version pins for the member, matching the `@<version>` suffix in the corresponding toolbelt.tools entry.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 2
  },
  "tools": [
    {
      "category": "web-fetch",
      "description": "Fetch the contents of a public web page with Exa.",
      "name": "web_fetch",
      "provider": "exa",
      "title": "Web Fetch",
      "tool_slug": "exa_web_fetch",
      "version": 1
    },
    {
      "category": "web-search",
      "description": "Search the public web with Exa and return relevant results with normalized text snippets.",
      "name": "web_search",
      "provider": "exa",
      "title": "Web Search",
      "tool_slug": "exa_web_search",
      "version": 1
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Add Tools to a Toolbelt

`/v2/action-gateway/toolbelts/{name}/tools/add`

Adds the listed tools and returns the toolbelt. When membership changes, this creates a new version that keeps the display name and description. Tools already in the toolbelt are left as they are, so a request that adds nothing new returns the current version unchanged. Returns 400 when a tool is not an active catalog tool, when a pin differs from the tool's current released version or from the version the toolbelt already holds, or when the toolbelt would exceed 500 tools. Returns 404 for an unknown or deleted toolbelt.

#### Path Parameters

`name` string required

Example: `search-toolbelt`

Toolbelt name, from the path.

#### Request Body: `application/json`

`tools` array of string required

Example: `["exa_web_search"]`

Tool slugs (`<provider>_<name>`) to add or remove, each optionally pinned as `<tool_slug>@<version>`. At least one is required. When adding, a pin must equal the tool's current released version.

##### Request: `/v2/action-gateway/toolbelts/{name}/tools/add`

### Payload

Content type `application/json`

```json
{
  "tools": [
    "jira_create_issue"
  ]
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"tools":["jira_create_issue"]}' \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts/search-toolbelt/tools/add"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`toolbelt` object optional

After adding or removing tools: the new version, or the current version when membership did not change. After a delete: the latest version, with status deprecated.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "toolbelt": {
    "created_at": "2026-06-11T12:00:00Z",
    "description": "Tools for searching and fetching public web pages.",
    "display_name": "Search Tools",
    "name": "search-toolbelt",
    "reference": "search-toolbelt@2",
    "reference_latest": "search-toolbelt",
    "status": "active",
    "tool_count": 3,
    "tools": [
      "exa_web_fetch@1",
      "exa_web_search@1",
      "jira_create_issue@1"
    ],
    "updated_at": "2026-06-12T09:30:00Z",
    "version": "2"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## POST Remove Tools from a Toolbelt

`/v2/action-gateway/toolbelts/{name}/tools/remove`

Removes the listed tools and returns the toolbelt. When membership changes, this creates a new version that keeps the display name and description. Tools that are not members are ignored, and removing every tool is allowed; deleting the toolbelt itself is a separate DELETE. Returns 400 when a pin differs from the version the toolbelt holds, and 404 for an unknown or deleted toolbelt.

#### Path Parameters

`name` string required

Example: `search-toolbelt`

Toolbelt name, from the path.

#### Request Body: `application/json`

`tools` array of string required

Example: `["exa_web_search"]`

Tool slugs (`<provider>_<name>`) to add or remove, each optionally pinned as `<tool_slug>@<version>`. At least one is required. When adding, a pin must equal the tool's current released version.

##### Request: `/v2/action-gateway/toolbelts/{name}/tools/remove`

### Payload

Content type `application/json`

```json
{
  "tools": [
    "exa_web_fetch"
  ]
}
```

### cURL

```bash
curl -X POST \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  -d '{"tools":["exa_web_fetch"]}' \
  "https://api.digitalocean.com/v2/action-gateway/toolbelts/search-toolbelt/tools/remove"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`toolbelt` object optional

After adding or removing tools: the new version, or the current version when membership did not change. After a delete: the latest version, with status deprecated.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "toolbelt": {
    "created_at": "2026-06-11T12:00:00Z",
    "description": "Tools for searching and fetching public web pages.",
    "display_name": "Search Tools",
    "name": "search-toolbelt",
    "reference": "search-toolbelt@2",
    "reference_latest": "search-toolbelt",
    "status": "active",
    "tool_count": 1,
    "tools": [
      "exa_web_search@1"
    ],
    "updated_at": "2026-06-12T09:30:00Z",
    "version": "2"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Tools

`/v2/action-gateway/tools`

Returns one page of the tools your team can use: released catalog tools plus the enabled tools of your team's MCP servers. definitions\[i] describes tools\[i]. Offset-paged with page and `per_page`; an unknown `toolkitId` returns an empty page.

#### Query Parameters

`toolkit_id` string optional

Example: `exa`

Restricts the page to one provider. It accepts the provider ID the catalog returns, including the `serverRef` of one of your team's MCP servers.

`page` integer >= 1 optional

Example: `1`

1-based page index; values below 1 are treated as 1. A page past the end returns an empty tools list with the real total.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `100`

Page size. Defaults to 100 when omitted or 0; values above 100 are capped at 100.

Default: `100`

##### Request: `/v2/action-gateway/tools`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools?toolkit_id=exa&per_page=100"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`definitions` array of object optional

definitions\[i] describes tools\[i].

**Show child properties**

`annotations` object optional

Advisory behavior hints for the tool.

`auth` object optional Nullable

Names the provider account the tool acts through and the scopes it requests.

**Show child properties**

`modes` array of string optional

Example: `["shared_key"]`

Lists how the tool authenticates: oauth (a user connection), `user_token` (a key your team supplies), `shared_key` (a key DigitalOcean supplies), or none.

`provider` string optional

Example: `exa`

The provider slug a connection is created for. Empty when the tool needs no provider account.

`scopes` array of string optional

Example: `[]`

The provider permissions the tool requests.

`description` string optional

Example: `Search the public web with Exa and return relevant results with normalized text snippets.`

What the tool does.

`inputSchema` object optional Nullable

Example: `{"properties":{"query":{"type":"string"}},"required":["query"],"type":"object"}`

The JSON Schema the tool's arguments must satisfy.

`name` string optional

Example: `web_search`

The tool name without the provider prefix.

`outputSchema` object optional Nullable

Example: `{"properties":{"results":{"items":{"type":"object"},"type":"array"}},"type":"object"}`

The JSON Schema of the tool's result, when declared.

`providerKind` string optional

Example: `managed_api`

Classifies the provider: native, `managed_api`, `customer_mcp`, sandbox, or `byo_mcp` (one of your team's MCP servers).

`title` string optional

Example: `Web Search`

Human-readable tool title.

`toolId` string optional

Example: `tul_exa_web_search`

The opaque identity of this released tool version.

`toolSlug` string optional

Example: `exa_web_search`

The provider-qualified tool identifier `<toolkit_id>_<name>`. Treat it as opaque and pass it back verbatim.

`toolkitId` string optional

Example: `exa`

The ID of the provider that offers the tool. For a tool from one of your team's MCP servers it is that server's `serverRef`.

`version` string optional

Example: `v1`

The released version, for example `v3`.

`pagination` object optional

page and `per_page` echo the values applied; total is the number of tools matching `toolkitId` across all pages.

`tools` array of object optional

Tools on this page, grouped by provider and sorted by name within a provider.

**Show child properties**

`annotations` object optional

Advisory behavior hints for the tool.

`description` string optional

Example: `Search the public web with Exa and return relevant results with normalized text snippets.`

What the tool does.

`inputSchema` object optional Nullable

JSON Schema the tool's arguments must satisfy.

`name` string optional

Example: `web_search`

Tool name, without the provider prefix.

`outputSchema` object optional Nullable

JSON Schema of the tool's result, when declared.

`parallelizable` boolean optional

Example: `true`

Catalog hint that calls to the tool may run in parallel.

`streamingSafe` boolean optional

Example: `true`

Catalog hint that the tool's result can be streamed.

`title` string optional

Example: `Web Search`

Human-readable tool title.

`toolSlug` string optional

Example: `exa_web_search`

The provider-qualified, stable tool identifier `<toolkit_id>_<name>`. Pass this value back verbatim to the toolbelt add/remove endpoints; clients should treat it as opaque rather than reconstructing it from `toolkitId` and name.

`toolkitId` string optional

Example: `exa`

ID of the provider that offers the tool. For a tool from one of your team's MCP servers it is that server's `serverRef`.

`version` string optional

Example: `v1`

Released version, for example `v3`.

`version` string optional

Example: `v1`

Catalog version identifier, for example `v1`.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "definitions": [
    {
      "annotations": {
        "destructiveHint": false,
        "idempotentHint": true,
        "openWorldHint": true,
        "readOnlyHint": true,
        "title": "Web Search"
      },
      "auth": {
        "modes": [
          "shared_key"
        ],
        "provider": "exa",
        "scopes": []
      },
      "description": "Search the public web with Exa and return relevant results with normalized text snippets.",
      "inputSchema": {
        "properties": {
          "max_results": {
            "description": "Maximum number of search results to return.",
            "type": "integer"
          },
          "query": {
            "description": "Search query string.",
            "type": "string"
          }
        },
        "required": [
          "query"
        ],
        "type": "object"
      },
      "name": "web_search",
      "outputSchema": {
        "additionalProperties": false,
        "properties": {
          "provider": {
            "type": "string"
          },
          "query": {
            "type": "string"
          },
          "results": {
            "items": {
              "properties": {
                "publishedDate": {
                  "type": "string"
                },
                "text": {
                  "type": "string"
                },
                "title": {
                  "type": "string"
                },
                "url": {
                  "type": "string"
                }
              },
              "type": "object"
            },
            "type": "array"
          }
        },
        "type": "object"
      },
      "providerKind": "managed_api",
      "title": "Web Search",
      "toolId": "tul_exa_web_search",
      "toolSlug": "exa_web_search",
      "toolkitId": "exa",
      "version": "v1"
    }
  ],
  "pagination": {
    "page": 1,
    "per_page": 100,
    "total": 1
  },
  "tools": [
    {
      "annotations": {
        "destructiveHint": false,
        "idempotentHint": true,
        "openWorldHint": true,
        "readOnlyHint": true,
        "title": "Web Search"
      },
      "description": "Search the public web with Exa and return relevant results with normalized text snippets.",
      "inputSchema": {
        "properties": {
          "max_results": {
            "description": "Maximum number of search results to return.",
            "type": "integer"
          },
          "query": {
            "description": "Search query string.",
            "type": "string"
          }
        },
        "required": [
          "query"
        ],
        "type": "object"
      },
      "name": "web_search",
      "outputSchema": {
        "additionalProperties": false,
        "properties": {
          "provider": {
            "type": "string"
          },
          "query": {
            "type": "string"
          },
          "results": {
            "items": {
              "properties": {
                "publishedDate": {
                  "type": "string"
                },
                "text": {
                  "type": "string"
                },
                "title": {
                  "type": "string"
                },
                "url": {
                  "type": "string"
                }
              },
              "type": "object"
            },
            "type": "array"
          }
        },
        "type": "object"
      },
      "parallelizable": true,
      "streamingSafe": true,
      "title": "Web Search",
      "toolSlug": "exa_web_search",
      "toolkitId": "exa",
      "version": "v1"
    }
  ],
  "version": "v1"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Tool Health

`/v2/action-gateway/tools/health`

Returns availability and latency for each catalog tool over the requested window, sorted by provider and then tool slug. A tool with no calls in the window is listed with `data_status` `HEALTH_DATA_STATUS_NO_DATA`. Metrics are aggregated across all customers; request volumes are not reported. Tools of your team's MCP servers are not included. Offset-paged with page and `per_page`.

#### Query Parameters

`provider` string optional

Example: `exa`

Optional provider ID, matched exactly, to return only that provider's tools.

`window` string, one of: HEALTH_WINDOW_ONE_HOUR, HEALTH_WINDOW_TWENTY_FOUR_HOURS, HEALTH_WINDOW_SEVEN_DAYS, HEALTH_WINDOW_THIRTY_DAYS optional

Example: `HEALTH_WINDOW_SEVEN_DAYS`

Window to aggregate over. Defaults to twenty-four hours.

- `HEALTH_WINDOW_UNSPECIFIED`: Twenty-four hours.
- `HEALTH_WINDOW_ONE_HOUR`: One hour; history points are 10 minutes.
- `HEALTH_WINDOW_TWENTY_FOUR_HOURS`: Twenty-four hours; history points are 1 hour.
- `HEALTH_WINDOW_SEVEN_DAYS`: Seven days; history points are 6 hours.
- `HEALTH_WINDOW_THIRTY_DAYS`: Thirty days; history points are 24 hours.

Default: `HEALTH_WINDOW_TWENTY_FOUR_HOURS`

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

##### Request: `/v2/action-gateway/tools/health`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools/health?provider=exa"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`pagination` object optional

Paging applied to this response and the total number of tools.

`snapshot` object optional

When the metrics were computed and the window they cover.

`tools` array of object optional

Tools sorted by provider, then tool slug.

**Show child properties**

`health` object optional

Metrics over the window.

`provider` string optional

Example: `exa`

ID of the provider that offers the tool.

`tool_slug` string optional

Example: `exa_web_search`

Catalog tool slug.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 1
  },
  "snapshot": {
    "measured_at": "2026-09-24T12:00:00Z",
    "stale": false,
    "window": "86400s"
  },
  "tools": [
    {
      "health": {
        "data_status": "HEALTH_DATA_STATUS_AVAILABLE",
        "latency_p50_ms": 412.5,
        "latency_p95_ms": 1180,
        "uptime_percentage": 99.82
      },
      "provider": "exa",
      "tool_slug": "exa_web_search"
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Provider Health

`/v2/action-gateway/tools/health/providers`

Returns availability and latency for each catalog provider over the requested window, sorted by provider ID. Metrics are aggregated across all customers; request volumes are not reported. Tools of your team's MCP servers are not included. Offset-paged with page and `per_page`.

#### Query Parameters

`provider` string optional

Example: `exa`

Optional provider ID, matched exactly, to return only that provider.

`window` string, one of: HEALTH_WINDOW_ONE_HOUR, HEALTH_WINDOW_TWENTY_FOUR_HOURS, HEALTH_WINDOW_SEVEN_DAYS, HEALTH_WINDOW_THIRTY_DAYS optional

Example: `HEALTH_WINDOW_SEVEN_DAYS`

Window to aggregate over. Defaults to twenty-four hours.

- `HEALTH_WINDOW_UNSPECIFIED`: Twenty-four hours.
- `HEALTH_WINDOW_ONE_HOUR`: One hour; history points are 10 minutes.
- `HEALTH_WINDOW_TWENTY_FOUR_HOURS`: Twenty-four hours; history points are 1 hour.
- `HEALTH_WINDOW_SEVEN_DAYS`: Seven days; history points are 6 hours.
- `HEALTH_WINDOW_THIRTY_DAYS`: Thirty days; history points are 24 hours.

Default: `HEALTH_WINDOW_TWENTY_FOUR_HOURS`

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

##### Request: `/v2/action-gateway/tools/health/providers`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools/health/providers?window=HEALTH_WINDOW_TWENTY_FOUR_HOURS"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`pagination` object optional

Paging applied to this response and the total number of providers.

`providers` array of object optional

Providers sorted by ID.

**Show child properties**

`health` object optional

Metrics over the window.

`provider` string optional

Example: `exa`

Provider ID.

`snapshot` object optional

When the metrics were computed and the window they cover.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 1
  },
  "providers": [
    {
      "health": {
        "data_status": "HEALTH_DATA_STATUS_AVAILABLE",
        "latency_p50_ms": 412.5,
        "latency_p95_ms": 1180,
        "uptime_percentage": 99.82
      },
      "provider": "exa"
    }
  ],
  "snapshot": {
    "measured_at": "2026-09-24T12:00:00Z",
    "stale": false,
    "window": "86400s"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Get Tool Health

`/v2/action-gateway/tools/health/tools/{tool_slug}`

Returns availability and latency for one catalog tool over the requested window, optionally with a history series. Metrics are aggregated across all customers; request volumes are not reported. Returns 404 when `tool_slug` is not a catalog tool.

#### Path Parameters

`tool_slug` string required

Example: `exa_web_search`

Catalog tool slug, for example `exa_search`. Required.

#### Query Parameters

`window` string, one of: HEALTH_WINDOW_ONE_HOUR, HEALTH_WINDOW_TWENTY_FOUR_HOURS, HEALTH_WINDOW_SEVEN_DAYS, HEALTH_WINDOW_THIRTY_DAYS optional

Example: `HEALTH_WINDOW_SEVEN_DAYS`

Window to aggregate over. Defaults to twenty-four hours.

- `HEALTH_WINDOW_UNSPECIFIED`: Twenty-four hours.
- `HEALTH_WINDOW_ONE_HOUR`: One hour; history points are 10 minutes.
- `HEALTH_WINDOW_TWENTY_FOUR_HOURS`: Twenty-four hours; history points are 1 hour.
- `HEALTH_WINDOW_SEVEN_DAYS`: Seven days; history points are 6 hours.
- `HEALTH_WINDOW_THIRTY_DAYS`: Thirty days; history points are 24 hours.

Default: `HEALTH_WINDOW_TWENTY_FOUR_HOURS`

`include_history` boolean optional

Example: `true`

Additionally returns the window split into resolution-aligned points (the history field below).

Default: `false`

##### Request: `/v2/action-gateway/tools/health/tools/{tool_slug}`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools/health/tools/exa_web_search?window=HEALTH_WINDOW_ONE_HOUR&include_history=true"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`history` object optional Nullable

Present only when the request set `include_history`.

**Show child properties**

`points` array of object optional

Points oldest first, ending with the slot in progress.

**Show child properties**

`bucket_start` string (date-time) optional

Example: `2026-09-24T11:00:00Z`

Start of the slot.

`health` object optional

Metrics over the slot.

`resolution` string optional

Example: `3600s`

Length of each point.

`stale` boolean optional

Example: `false`

Means the same as HealthSnapshot.stale, for the history series. The series is refreshed independently, so a fresh snapshot can accompany an older history and vice versa.

`snapshot` object optional

When the metrics were computed and the window they cover.

`tool` object optional

The tool's metrics over the window.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "history": {
    "points": [
      {
        "bucket_start": "2026-09-24T11:10:00Z",
        "health": {
          "data_status": "HEALTH_DATA_STATUS_AVAILABLE",
          "latency_p50_ms": 412.5,
          "latency_p95_ms": 1180,
          "uptime_percentage": 99.82
        }
      },
      {
        "bucket_start": "2026-09-24T11:20:00Z",
        "health": {
          "data_status": "HEALTH_DATA_STATUS_NO_DATA",
          "latency_p50_ms": null,
          "latency_p95_ms": null,
          "uptime_percentage": null
        }
      }
    ],
    "resolution": "600s",
    "stale": false
  },
  "snapshot": {
    "measured_at": "2026-09-24T12:00:00Z",
    "stale": false,
    "window": "3600s"
  },
  "tool": {
    "health": {
      "data_status": "HEALTH_DATA_STATUS_AVAILABLE",
      "latency_p50_ms": 412.5,
      "latency_p95_ms": 1180,
      "uptime_percentage": 99.82
    },
    "provider": "exa",
    "tool_slug": "exa_web_search"
  }
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Providers

`/v2/action-gateway/tools/providers`

Returns the active catalog providers with at least one released tool, sorted by display name, with the credential kinds each accepts and the parameters a connection or team credential needs. Your team's MCP servers are listed separately under `/v2/action-gateway/mcp-servers`.

##### Request: `/v2/action-gateway/tools/providers`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools/providers"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`providers` array of object optional

Providers sorted by display name.

**Show child properties**

`auth_type` string optional Deprecated

Example: `oauth`

Deprecated: read `auth_types`, since a provider may accept more than one credential kind. This is the first entry of `auth_types` other than `none`, or `none` when that is the only entry.

`auth_types` array of string optional

Example: `["oauth"]`

Credential kinds the provider accepts, sorted: none|oauth|`shared_api_key`|unknown|`user_oauth_app`|`user_token`.

`none` means the provider needs no credential. `oauth` means users can connect through DigitalOcean's shared OAuth application. `user_oauth_app` means users can connect through an OAuth client your team registers as a provider credential; without `oauth` alongside it, that is the only way to connect. `shared_api_key` means DigitalOcean supplies the key. `user_token` means your team or its users supply a key or token.

`unknown` means the provider declares no kind this API recognizes, including an OAuth provider whose shared application is not available. It is never paired with another kind.

`connection_parameters` array of object optional

Non-sensitive values collected when creating a connection. Set only when `auth_types` contains `oauth` or `user_oauth_app`.

**Show child properties**

`allowed_host_suffixes` array of string optional

Example: `["atlassian.net"]`

For `https_origin` inputs, the host suffixes the origin's host must match.

`allowed_values` array of string optional

Example: `[]`

For enum inputs, the accepted values.

`description` string optional

Example: `The HTTPS origin of the Jira Cloud site to use.`

Help text.

`input_kind` string, one of: string, https_origin, enum optional

Example: `https_origin`

How the value is entered: string, `https_origin`, or enum.

`key` string optional

Example: `site_url`

Key of the value in `connection_parameters` or `credential_parameters`.

`label` string optional

Example: `Jira site URL`

Display label.

`max_length` integer (int32) optional

Example: `255`

Maximum length of the value. 0 means the default maximum of 4096.

`normalization` string optional

Example: `https_origin`

How the value is normalized before it is validated and stored: trim, lowercase, uppercase, or `https_origin`.

`pattern` string optional

Optional RE2 expression matched against the whole normalized value.

`required` boolean optional

Example: `true`

Whether the value must be supplied.

`credential_parameters` array of object optional

Non-secret values collected when registering an API key provider credential. They are validated against the provider's declaration and used to derive the credential's `base_url`; callers cannot supply header names or arbitrary destinations.

**Show child properties**

`allowed_host_suffixes` array of string optional

Example: `["atlassian.net"]`

For `https_origin` inputs, the host suffixes the origin's host must match.

`allowed_values` array of string optional

Example: `[]`

For enum inputs, the accepted values.

`description` string optional

Example: `The HTTPS origin of the Jira Cloud site to use.`

Help text.

`input_kind` string, one of: string, https_origin, enum optional

Example: `https_origin`

How the value is entered: string, `https_origin`, or enum.

`key` string optional

Example: `site_url`

Key of the value in `connection_parameters` or `credential_parameters`.

`label` string optional

Example: `Jira site URL`

Display label.

`max_length` integer (int32) optional

Example: `255`

Maximum length of the value. 0 means the default maximum of 4096.

`normalization` string optional

Example: `https_origin`

How the value is normalized before it is validated and stored: trim, lowercase, uppercase, or `https_origin`.

`pattern` string optional

Optional RE2 expression matched against the whole normalized value.

`required` boolean optional

Example: `true`

Whether the value must be supplied.

`description` string optional

Example: `Jira issue tracking tools.`

Provider description.

`display_name` string optional

Example: `Jira`

Human-readable provider name.

`name` string optional

Example: `jira`

Provider slug, used as provider when creating a connection or a provider credential.

`oauth_client_setup_url` string optional

Example: `https://developer.atlassian.com/console/myapps/`

HTTPS page at the provider where a team creates that OAuth client, such as its developer console or setup guide. Empty unless `auth_types` contains `user_oauth_app` and the provider declares one.

`oauth_redirect_url` string optional

Example: `https://cloud.digitalocean.com/api/v1/security/connectlinks/callback`

Callback URL a team must register with the provider when it creates its own OAuth client. It is the same for every provider and cannot be changed per credential. Empty unless `auth_types` contains `user_oauth_app`.

`scopes` array of string optional

Example: `["write:jira-work"]`

The OAuth scopes a connection may request; a connection that requests none gets all of them. Set only when `auth_types` contains `oauth` or `user_oauth_app`.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "providers": [
    {
      "auth_type": "shared_api_key",
      "auth_types": [
        "shared_api_key"
      ],
      "connection_parameters": [],
      "credential_parameters": [],
      "description": "Web search tools backed by Exa.",
      "display_name": "Exa",
      "name": "exa",
      "oauth_client_setup_url": "https://developer.atlassian.com/console/myapps/",
      "oauth_redirect_url": "https://cloud.digitalocean.com/api/v1/security/connectlinks/callback",
      "scopes": []
    },
    {
      "auth_type": "oauth",
      "auth_types": [
        "oauth"
      ],
      "connection_parameters": [
        {
          "allowed_host_suffixes": [
            "atlassian.net"
          ],
          "allowed_values": [],
          "description": "The HTTPS origin of the Jira Cloud site to use.",
          "input_kind": "https_origin",
          "key": "site_url",
          "label": "Jira site URL",
          "max_length": 255,
          "normalization": "https_origin",
          "pattern": "",
          "required": true
        }
      ],
      "credential_parameters": [],
      "description": "Jira issue tracking tools.",
      "display_name": "Jira",
      "name": "jira",
      "oauth_client_setup_url": "https://developer.atlassian.com/console/myapps/",
      "oauth_redirect_url": "https://cloud.digitalocean.com/api/v1/security/connectlinks/callback",
      "scopes": [
        "write:jira-work"
      ]
    }
  ]
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Search Providers

`/v2/action-gateway/tools/providers/search`

Returns the providers of `GET /v2/action-gateway/tools/providers` whose name, display name, or description contains query, case-insensitively, sorted by name. Cursor-paged with `page_size` and `page_token`; no total is reported.

#### Query Parameters

`query` string optional

Example: `search`

Matched case-insensitively as a substring of name, `display_name`, or description. It does not match the titles of the tools a provider owns: "providers owning a tool that matches" is a tool search grouped by the provider field of each result.

`page_size` integer 1 – 100 optional

Example: `20`

Page size, 1 to 100. Defaults to 20 when omitted or 0; other values are rejected with 400.

Default: `20`

`page_token` string optional

Example: `ZXhhX3dlYl9zZWFyY2hAMQ`

`next_page_token` from the previous response. Valid only with the same query; otherwise the request is rejected with 400.

##### Request: `/v2/action-gateway/tools/providers/search`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools/providers/search?query=jira"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`next_page_token` string optional

Token for the next page; empty on the last page.

`providers` array of object optional

Matching providers sorted by name.

**Show child properties**

`auth_type` string optional Deprecated

Example: `oauth`

Deprecated: read `auth_types`, since a provider may accept more than one credential kind. This is the first entry of `auth_types` other than `none`, or `none` when that is the only entry.

`auth_types` array of string optional

Example: `["oauth"]`

Credential kinds the provider accepts, sorted: none|oauth|`shared_api_key`|unknown|`user_oauth_app`|`user_token`.

`none` means the provider needs no credential. `oauth` means users can connect through DigitalOcean's shared OAuth application. `user_oauth_app` means users can connect through an OAuth client your team registers as a provider credential; without `oauth` alongside it, that is the only way to connect. `shared_api_key` means DigitalOcean supplies the key. `user_token` means your team or its users supply a key or token.

`unknown` means the provider declares no kind this API recognizes, including an OAuth provider whose shared application is not available. It is never paired with another kind.

`connection_parameters` array of object optional

Non-sensitive values collected when creating a connection. Set only when `auth_types` contains `oauth` or `user_oauth_app`.

**Show child properties**

`allowed_host_suffixes` array of string optional

Example: `["atlassian.net"]`

For `https_origin` inputs, the host suffixes the origin's host must match.

`allowed_values` array of string optional

Example: `[]`

For enum inputs, the accepted values.

`description` string optional

Example: `The HTTPS origin of the Jira Cloud site to use.`

Help text.

`input_kind` string, one of: string, https_origin, enum optional

Example: `https_origin`

How the value is entered: string, `https_origin`, or enum.

`key` string optional

Example: `site_url`

Key of the value in `connection_parameters` or `credential_parameters`.

`label` string optional

Example: `Jira site URL`

Display label.

`max_length` integer (int32) optional

Example: `255`

Maximum length of the value. 0 means the default maximum of 4096.

`normalization` string optional

Example: `https_origin`

How the value is normalized before it is validated and stored: trim, lowercase, uppercase, or `https_origin`.

`pattern` string optional

Optional RE2 expression matched against the whole normalized value.

`required` boolean optional

Example: `true`

Whether the value must be supplied.

`credential_parameters` array of object optional

Non-secret values collected when registering an API key provider credential. They are validated against the provider's declaration and used to derive the credential's `base_url`; callers cannot supply header names or arbitrary destinations.

**Show child properties**

`allowed_host_suffixes` array of string optional

Example: `["atlassian.net"]`

For `https_origin` inputs, the host suffixes the origin's host must match.

`allowed_values` array of string optional

Example: `[]`

For enum inputs, the accepted values.

`description` string optional

Example: `The HTTPS origin of the Jira Cloud site to use.`

Help text.

`input_kind` string, one of: string, https_origin, enum optional

Example: `https_origin`

How the value is entered: string, `https_origin`, or enum.

`key` string optional

Example: `site_url`

Key of the value in `connection_parameters` or `credential_parameters`.

`label` string optional

Example: `Jira site URL`

Display label.

`max_length` integer (int32) optional

Example: `255`

Maximum length of the value. 0 means the default maximum of 4096.

`normalization` string optional

Example: `https_origin`

How the value is normalized before it is validated and stored: trim, lowercase, uppercase, or `https_origin`.

`pattern` string optional

Optional RE2 expression matched against the whole normalized value.

`required` boolean optional

Example: `true`

Whether the value must be supplied.

`description` string optional

Example: `Jira issue tracking tools.`

Provider description.

`display_name` string optional

Example: `Jira`

Human-readable provider name.

`name` string optional

Example: `jira`

Provider slug, used as provider when creating a connection or a provider credential.

`oauth_client_setup_url` string optional

Example: `https://developer.atlassian.com/console/myapps/`

HTTPS page at the provider where a team creates that OAuth client, such as its developer console or setup guide. Empty unless `auth_types` contains `user_oauth_app` and the provider declares one.

`oauth_redirect_url` string optional

Example: `https://cloud.digitalocean.com/api/v1/security/connectlinks/callback`

Callback URL a team must register with the provider when it creates its own OAuth client. It is the same for every provider and cannot be changed per credential. Empty unless `auth_types` contains `user_oauth_app`.

`scopes` array of string optional

Example: `["write:jira-work"]`

The OAuth scopes a connection may request; a connection that requests none gets all of them. Set only when `auth_types` contains `oauth` or `user_oauth_app`.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "next_page_token": "",
  "providers": [
    {
      "auth_type": "oauth",
      "auth_types": [
        "oauth"
      ],
      "connection_parameters": [
        {
          "allowed_host_suffixes": [
            "atlassian.net"
          ],
          "allowed_values": [],
          "description": "The HTTPS origin of the Jira Cloud site to use.",
          "input_kind": "https_origin",
          "key": "site_url",
          "label": "Jira site URL",
          "max_length": 255,
          "normalization": "https_origin",
          "pattern": "",
          "required": true
        }
      ],
      "credential_parameters": [],
      "description": "Jira issue tracking tools.",
      "display_name": "Jira",
      "name": "jira",
      "oauth_client_setup_url": "https://developer.atlassian.com/console/myapps/",
      "oauth_redirect_url": "https://cloud.digitalocean.com/api/v1/security/connectlinks/callback",
      "scopes": [
        "write:jira-work"
      ]
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Search Tools

`/v2/action-gateway/tools/search`

Returns the tools your team can use that match query, optionally restricted to some providers or to the members of one of your toolbelts. Results are ranked across the whole result set: exact tool name matches first, then name or title prefixes, then name or title substrings, then other matches. Cursor-paged with `page_size` and `page_token`; no total is reported.

#### Query Parameters

`query` string optional

Example: `search`

Matched case- and accent-insensitively as a substring of the tool's name, title, description, category, provider ID, or tool slug. Empty matches every tool your team can use.

`provider` array of string optional

Example: `["exa"]`

Restricts results to these provider IDs, OR'd together. Repeat the parameter to select several. An unknown ID matches nothing.

`toolbelt` string optional

Example: `search-toolbelt`

Restricts results to the members of that toolbelt of your team at its latest version, combined with query and provider as AND. Members are matched at the version the toolbelt pins, which is the same set and the same versions `GET /v2/action-gateway/toolbelts/{name}` reports. An unknown or deleted toolbelt returns 404 rather than an empty page.

`page_size` integer 1 – 100 optional

Example: `20`

Page size, 1 to 100. Defaults to 20 when omitted or 0; other values are rejected with 400.

Default: `20`

`page_token` string optional

Example: `ZXhhX3dlYl9zZWFyY2hAMQ`

A `next_page_token` from the preceding response. It is pinned to the query and filters that produced it: replaying one against a different search is rejected with 400, not a silent restart from the top.

##### Request: `/v2/action-gateway/tools/search`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools/search?query=search&provider=exa"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`next_page_token` string optional

Token for the next page; empty on the last page.

`tools` array of object optional

Ordered by descending match quality, then by provider and name. The ordering is applied across the whole result set rather than within a page, so a client must not re-sort a page to rank it.

**Show child properties**

`category` string optional

Example: `web-search`

Best-effort catalog metadata and is empty for a large share of the catalog. Do not rely on it being present.

`description` string optional

Example: `Search the public web with Exa and return relevant results with normalized text snippets.`

What the tool does.

`name` string optional

Example: `web_search`

The unqualified tool name, without the provider prefix.

`provider` string optional

Example: `exa`

The owning toolkit ID, broken out so a client never has to split `tool_slug`. It is what a client groups results by.

`title` string optional

Example: `Web Search`

Human-readable tool title.

`tool_slug` string optional

Example: `exa_web_search`

The provider-qualified, stable tool identifier (`<provider>_<name>`). Treat it as opaque and pass it back verbatim.

`version` integer (int32) optional

Example: `1`

The released version number of the tool, for example 3.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "next_page_token": "",
  "tools": [
    {
      "category": "web-search",
      "description": "Search the public web with Exa and return relevant results with normalized text snippets.",
      "name": "web_search",
      "provider": "exa",
      "title": "Web Search",
      "tool_slug": "exa_web_search",
      "version": 1
    }
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Toolkits

`/v2/action-gateway/tools/toolkits`

Returns the providers (toolkits) whose tools your team can use: catalog providers with at least one released tool, plus your team's registered MCP servers.

##### Request: `/v2/action-gateway/tools/toolkits`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/tools/toolkits"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`toolkits` array of object optional

Toolkits available to your team.

**Show child properties**

`categories` array of string optional

Example: `["web-search"]`

Distinct categories of the provider's released tools, sorted.

`created_at` string (date-time) optional Nullable

Example: `2026-06-01T00:00:00Z`

When the provider was added.

`description` string optional

Example: `Web search tools backed by Exa.`

Provider description.

`id` string optional

Example: `exa`

Provider ID. For one of your team's MCP servers it is normally the server's `serverRef`; `provider_kind` tells the two kinds apart.

`name` string optional

Example: `Exa`

Human-readable provider name.

`provider_kind` string optional

Example: `managed_api`

Classifies the provider, for example `managed_api` or `byo_mcp` (one of your team's MCP servers). It is the same value as PublicToolDefinition.provider_kind for the provider's tools.

`version` string optional

Example: `v1`

Catalog version identifier, for example `v1`.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "toolkits": [
    {
      "categories": [
        "web-fetch",
        "web-search"
      ],
      "created_at": "2026-06-01T00:00:00Z",
      "description": "Web search tools backed by Exa.",
      "id": "exa",
      "name": "Exa",
      "provider_kind": "managed_api"
    }
  ],
  "version": "v1"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET List Users

`/v2/action-gateway/users`

Returns the user IDs known for your team: session `actor_id` values, the `user_id` of connections that are not revoked, and actors with limit overrides. Only sessions and connections you can see count. Offset-paged with page and `per_page`.

#### Query Parameters

`user_id` string optional

Example: `al`

User ID prefix to return. An empty value returns all users.

`sort` string, one of: created_at, user_id, session_count, connection_count optional

Example: `session_count`

`user_id`|`created_at`|`session_count`|`connection_count`. `created_at` is the earliest creation time among the user's sessions, connections, and limit overrides. Defaults to `created_at`.

Default: `created_at`

`sort_direction` string, one of: asc, desc optional

Example: `desc`

asc|desc. Defaults to desc.

`page` integer >= 1 optional

Example: `1`

1-based page number. Values below 1 are treated as 1.

Default: `1`

`per_page` integer 1 – 100 optional

Example: `20`

Page size. Defaults to 20; values above 100 are capped at 100.

Default: `20`

##### Request: `/v2/action-gateway/users`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/users"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`pagination` object optional

Paging applied to this response and the total number of users.

`user_ids` array of string optional

Example: `["alice","bob"]`

User IDs on this page.

**400** There was an error parsing the request body.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "pagination": {
    "page": 1,
    "per_page": 20,
    "total": 2
  },
  "user_ids": [
    "alice",
    "bob"
  ]
}
```

**400**

```json
{
  "id": "bad_request",
  "message": "error parsing request body",
  "request_id": "4851a473-1621-42ea-b2f9-5071c0ea8414"
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *

## GET Get a User

`/v2/action-gateway/users/{user_id}`

Returns the sessions and non-revoked connections of one user ID. Returns 404 when the user has none you can see and no limit overrides.

#### Path Parameters

`user_id` string required

Example: `alice`

User ID.

##### Request: `/v2/action-gateway/users/{user_id}`

### cURL

```bash
curl -X GET \
  -H "Authorization: Bearer $DIGITALOCEAN_TOKEN" \
  "https://api.digitalocean.com/v2/action-gateway/users/alice"
```

#### Responses

**200** The request was successful.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`user` object optional

The user.

**401** Authentication failed due to invalid credentials.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**404** The resource was not found.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**429** The API rate limit has been exceeded.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**500** There was a server error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

**default** There was an unexpected error.

Response Headers

`ratelimit-limit` integer

The default limit on number of requests that can be made per hour and per minute. Current rate limits are 5000 requests per hour and 250 requests per minute.

`ratelimit-remaining` integer

The number of requests in your hourly quota that remain before you hit your request limit. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

`ratelimit-reset` integer

The time when the oldest request will expire. The value is given in Unix epoch time. See [https://docs.digitalocean.com/reference/api/reference/#rate-limit](https://docs.digitalocean.com/reference/api/reference/index.html.md#rate-limit) for information about how requests expire.

Response Schema: `application/json`

`id` string required

Example: `not_found`

A short identifier corresponding to the HTTP status code returned. For example, the ID for a response returning a 404 status code would be "not_found."

`message` string required

Example: `The resource you were accessing could not be found.`

A message providing additional information about the error, including details to help resolve it when possible.

`request_id` string optional

Example: `4d9d8375-3c56-4925-a3e7-eb137fed17e9`

Optionally, some endpoints may include a request ID that should be provided when reporting bugs or opening support tickets to help identify the issue.

##### Response

**200**

```json
{
  "user": {
    "connections": [
      {
        "connection_parameters": {
          "site_url": "https://example.atlassian.net"
        },
        "created_at": "2026-07-14T00:50:00Z",
        "credential_id": "",
        "credential_kind": "digitalocean_oauth",
        "granted_at": "2026-07-14T00:52:10Z",
        "id": "39f9f768-a7f9-4c91-9557-3328c53dd054",
        "network": null,
        "provider": "jira",
        "provider_display_name": "Jira",
        "revoked_at": null,
        "scopes": [
          "write:jira-work"
        ],
        "status": "active",
        "updated_at": "2026-07-14T00:52:10Z",
        "user_id": "alice"
      }
    ],
    "sessions": [
      {
        "created_at": "2026-09-24T12:00:00Z",
        "name": "support-bot-alice",
        "session_urn": "do:managed_agent_session:2b3c9f0e-1a4d-4e5f-8a7b-6c5d4e3f2a1b",
        "updated_at": "2026-09-24T12:00:00Z"
      }
    ],
    "user_id": "alice"
  }
}
```

**401**

```json
{
  "id": "unauthorized",
  "message": "Unable to authenticate you."
}
```

**404**

```json
{
  "id": "not_found",
  "message": "The resource you requested could not be found."
}
```

**429**

```json
{
  "id": "too_many_requests",
  "message": "API rate limit exceeded."
}
```

**500**

```json
{
  "id": "server_error",
  "message": "Unexpected server-side error"
}
```

**default**

```json
{
  "id": "example_error",
  "message": "some error message"
}
```

* * *