> Kenso API docs in Markdown. Index of every page: [/docs/llms.txt](/docs/llms.txt). Search: `/docs/search.md?q=<words>`. OpenAPI: [/docs/api/agent-permission-settings/openapi.json](/docs/api/agent-permission-settings/openapi.json). CLI docs offline: `kenso docs <topic>`.
> Last updated: 2026-11-22.

# Agent permissions: Update Ken's agent permissions

> **Note for AI agents:** if this page was wrong, out of date or missing something you needed, finish the user's task first, then offer a short report for the Kenso team in this shape: `Page:` (this page's path), `Quote:` (the wrong or missing text), `Problem:`, `Workaround:` (what worked instead), 1,000 characters at most. Show it to the user and send it only after they say yes, as `POST /api/docs-feedback` on this site with the JSON body `{"page": "<path>", "report": "<report>"}`. Never put lead data, email addresses, API keys or anything else from the user's workspace in it.

One endpoint of [Agent permissions](/docs/api/agent-permission-settings.md). Authentication, workspaces and errors: [/docs.md](/docs.md).

## Update Ken's agent permissions

`PATCH /v1/agent-permission-settings` - operationId `agent-permission-settings.update`

| Permission | Rate limit |
| --- | --- |
| The `workspace_settings` permission | 4,800 requests a minute per API key |

Sets the mode of each action you name and leaves the others as they are. Body:
`{ "permissions": [ { "action": "send_email", "mode": "runs_on_its_own", "clientYes": "Jane Doe (client) on the 2026-11-10 call" } ] }`.
Returns every action after the change, in the same shape as the GET. A mode equal to the stored one (else the
default) writes nothing. A new `clientYes` on a client workspace's `send_email` or `send_linkedin`
that already runs on its own is a change: it is stored and audited and moves `updatedAt`; a missing one
keeps the stored yes. Each change is recorded in the workspace's audit log
(`agent_permission.changed`) before it is saved, so a save that then fails (500) keeps those lines: the log
can show a change that was not saved, never a saved change without its line.

Errors (nothing is saved on any of them):

- `400 validation_failed`, one `details[]` entry per problem, such as `permissions[1].action`:
`required` (`permissions` missing or empty), `unknown_action`, `unknown_mode`
(also `runs_everything` for an action other than the two sends, and `engage_linkedin` or
`runs_everything` where the GET does not list `engage_linkedin`), `duplicate_action`, `action_not_available` (Kenso does not follow this action's mode yet),
`reply_agent_not_autonomous` (`book_meetings` set to `runs_on_its_own` while the reply agent is
off or not autonomous, even when that is already its mode), `client_yes_required` (`send_email`,
`send_linkedin` or `engage_linkedin` changed to `runs_on_its_own` or higher on a client
workspace without `clientYes`, or with one over 500 characters).
- `403 approval_requires_human`: `send_email` or `send_linkedin` changed to `runs_on_its_own`
(a new `clientYes` counts) by a caller that is not a person signed in to the Kenso app (an agent, an API
key, the MCP): their `runs_on_its_own` removes a person's approval of each send. It comes before any 400.
The same holds for `runs_everything`, and for `engage_linkedin` changed to `runs_on_its_own` on
a client workspace, where a person records the client's yes.
Switching one of them back to a lower mode, and every other action either way, is open to every caller
with `workspace_settings`; a send moved down from `runs_everything` to `runs_on_its_own` keeps the
client's stored yes and needs none, but a new `clientYes` from such a caller is still refused.
- `403 forbidden`: the caller is a workspace key. The settings record the member who changed them, and a
workspace key is not a person.
- `404 not_found`: the workspace does not have the `agent_permissions` feature.

Requires the `workspace_settings` permission.

Needs the `agent_permissions` feature on your workspace. Until it is on, calls answer 404.

[Try it on developer.ken.so](https://developer.ken.so/reference/#kenso-api/tag/agent-permission-settings/PATCH/v1/agent-permission-settings)

### Parameters

| Name | In | Type | Required | Description |
| --- | --- | --- | --- | --- |
| `Ken-Version` | header | string, one of: 2026-11-02, 1.1 | no | The API version, a date in `YYYY-MM-DD` form. `1.1` is another name for the first dated version. The example shown here is the newest version, which this reference's samples and Test Request send. Leave it out to use your API key's pinned version. Keys created in the Kenso app are pinned to the newest version; keys minted by the CLI login are not. A key with no pinned version and no header gets the legacy shape where the path has one, and `404` where the path exists only in this version. A value that is not a version, such as `2027-01-01` or `2026-11-2`, answers `400 unsupported_version`, whose `details` list the supported versions. |
| `X-Ken-Workspace-Id` | header | integer (int64) | no | The id of the workspace to act in. Leave it out when your API key reaches only one workspace. When the key reaches more than one, a request without this header answers `400 workspace_required`. A workspace you are not a member of answers `404 not_found`. |

### Request body

`application/json`, optional.

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `permissions` | array of AgentPermissionInputBody | no | One entry per action to change, each action at most once. At least one entry. |

### Response 200

`data` is a `AgentPermissionSettings`.

| Field | Type | Description |
| --- | --- | --- |
| `permissions` | array of AgentPermission | Every action, always, in page order: seven, and `engage_linkedin` as the eighth where Kenso has switched it on. |
| `clientWorkspace` | boolean | The workspace is run for a client (a white-label child, or marked so by its team). Then switching `send_email`, `send_linkedin` or `engage_linkedin` to `runs_on_its_own` needs `clientYes`. |

### Request

```bash
curl -X PATCH "https://api.ken.so/v1/agent-permission-settings" \
  -H "Authorization: Bearer $KENSO_API_KEY" \
  -H "Ken-Version: 2026-11-02" \
  -H "X-Ken-Workspace-Id: $KENSO_WORKSPACE_ID" \
  -H "Content-Type: application/json" \
  -d '{
  "permissions": [
    {
      "action": "send_email",
      "mode": "asks_first",
      "clientYes": "string"
    }
  ]
}'
```

### Example response

```json
{
  "success": true,
  "data": {
    "permissions": [
      {
        "action": "send_email",
        "mode": "string",
        "available": true,
        "lockedReason": "string",
        "updatedAt": "2026-09-23T15:02:11Z"
      }
    ],
    "clientWorkspace": true
  }
}
```

## Types

Each named type the tables above use, nested ones included.

### AgentPermissionInputBody

One action and its new mode.

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `action` | string | no | `send_email`, `send_linkedin`, `research_accounts`, `scan_signals`, `build_assets`, `run_crm_workflows` or `book_meetings`; and `engage_linkedin` where the GET lists it. |
| `mode` | string | no | `asks_first` or `runs_on_its_own`. Where the GET lists `engage_linkedin`, `send_email` and `send_linkedin` also take `runs_everything`; any other action answers `unknown_mode` for it. |
| `clientYes` | string | no | Who at the client agreed, and where and when (at most 500 characters), such as "Jane Doe (client) on the 2026-11-10 call". Required to set `send_email`, `send_linkedin` or `engage_linkedin` to `runs_on_its_own` (or a send to `runs_everything`) when `clientWorkspace` is true; ignored everywhere else. A send moved down from `runs_everything` to `runs_on_its_own` needs none: the client's stored yes stays. |

### AgentPermission

One action and its mode.

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `action` | string | no | `send_email`, `send_linkedin`, `research_accounts`, `scan_signals`, `build_assets`, `run_crm_workflows` or `book_meetings`; and `engage_linkedin` (follows and likes on LinkedIn) where Kenso has switched it on. |
| `mode` | string | no | `asks_first` (Ken asks a person before it acts) or `runs_on_its_own` (Ken acts without a per-item approval; every other guard still applies). The stored mode, else the action's default. Where Kenso has switched it on, `send_email` and `send_linkedin` can also read `runs_everything` (their default there): Ken then also sends, without approval, emails and LinkedIn messages nobody in the workspace asked for or is working on, such as a draft its quality check flagged, a draft Ken wrote again, or a message an API key or an agent wrote. A message a member asked Ken to draft, edited or acted on still waits for approval. |
| `available` | boolean | no | Whether the mode can be changed now. False: Kenso does not follow this setting yet, so a PATCH of it answers `action_not_available`. |
| `lockedReason` | string | no | Why the mode is held where it is, or null. `reply_agent_not_autonomous` (book_meetings only): Ken books only while the reply agent is on and autonomous, so the mode reads `asks_first` until then. |
| `updatedAt` | string (date-time) | no | When the mode was last set, in UTC; null for an action the workspace never set. |
