Automette

The Automette REST API lets you render documents from your templates, retrieve renders, manage webhooks, and create Dynamic URLs.

Base URL

https://automette.com

All endpoints are versioned and live under /api/v1.

Conventions

Content type: Request and response bodies are JSON. Send Content-Type: application/json on every POST, PATCH, and DELETE.
Identifiers: resource IDs (templates, renders, jobs, webhook subscriptions, forms) are CUIDs, 25-character lowercase alphanumeric strings like cm4rnd7qh0001jx04w8e2t5va. Treat them as opaque. The one exception is event IDs in webhook payloads, which are prefixed evt_.
Timestamps: ISO 8601 UTC strings (2026-04-22T12:00:00.000Z).
self URL: List and single-resource responses include a self field with the canonical URL of the resource. Always prefer following self over reconstructing URLs by hand.

Authentication

Every request must include an Authorization header with a team-scoped API key:

Authorization: Bearer dg_your_api_key

See Authentication for full details on creating and rotating keys.

Errors

The API returns standard HTTP status codes and a JSON body shaped like:

{ "error": "Human-readable message", "detail": "Optional extra context" }

Code	When
`400`	Invalid JSON body or missing required field
`401`	Missing, malformed, or unrecognized API key
`402`	Insufficient credits for this operation
`404`	Resource does not exist or is not visible to your team (including resources owned by other teams)
`422`	Validation failed (e.g. unknown event, bad URL format)
`425`	Render not yet ready for download
`500`	Server error: safe to retry

Insufficient-credit errors (402) include extra fields so you can react programmatically:

{
  "error": "Insufficient credits",
  "required_credits": 1,
  "available_credits": 0,
  "credit_balance": 0,
  "overage_limit": 0
}

Rate limits

Each API key has two limits, counted per minute:

Limit	Applies to
600 requests	Every request with the key, including MCP
120 renders	`POST /api/v1/renders` and the preview endpoints

Every response carries X-RateLimit-Limit, X-RateLimit-Remaining and X-RateLimit-Reset (seconds until the minute ends). Over a limit, the request returns 429 with a Retry-After header:

{
  "error": "Rate limit exceeded: 120 renders per minute for this API key. Try again in 25 seconds.",
  "code": "rate_limited",
  "limit": 120,
  "retry_after": 25
}

Wait retry_after seconds and send it again. Rendering is also limited by your credit balance. To learn when a render finishes, use a webhook instead of polling. If you need higher limits, contact us.

Idempotency

POST /api/v1/renders is not idempotent: retrying a failed POST will create a new render. To avoid duplicate work after a network error, use async mode and rely on the webhook_url field for completion notification, then dedupe on render_id in your handler.

MCP

Everything below is also reachable over MCP, at https://automette.com/api/mcp. The tools there are task-shaped rather than one per endpoint, and an API key authenticates them exactly as it does here.

Two endpoints exist mainly for callers that are not us: GET /api/v1/skills and GET /api/v1/skills/{id} serve the template-authoring reference our own editor runs on, so a model working against the REST API is as well informed as one working in the app.

Versioning

/api/v1 is the only supported version. Breaking changes will be released under a new path (/api/v2) with a deprecation window for v1.

Last updated on

API Reference