HTTP 403Dangerous AI operations are disabled for this workspace

dangerous_ops_disabled

DELETE /v1/posts/{id} and POST /v1/posts/{id}/publish are guarded: they can delete content or publish publicly to real social accounts, so they stay off for API callers until a workspace owner explicitly opts in. This is the same toggle that gates MCP agents' publish/delete tools.

This page is where the problem type URI https://usepostify.com/docs/api/problems/dangerous-ops-disabled resolves.

Common causes

  • The workspace owner hasn't enabled "Dangerous AI operations" in Settings → API keys.
  • The toggle was switched back off after previously being enabled.

How to fix

  • Have an org owner open Settings → API keys and enable Dangerous AI operations.
  • Until then, create posts as drafts or scheduled posts (allowed) and publish from the app.
  • Design agent flows to propose-and-confirm: draft by API, let a human publish.

Example response

403 · application/problem+json
{
  "type": "https://usepostify.com/docs/api/problems/dangerous-ops-disabled",
  "title": "Dangerous AI operations are disabled for this workspace",
  "status": 403,
  "code": "dangerous_ops_disabled",
  "detail": "Dangerous operations are disabled for this workspace. An owner can enable them in Settings → API keys.",
  "request_id": "req_6f1f8a2b40d34cd2a6f1e2c9b8d7a601"
}

Related