Connect to the Revdoku API

Connect your own app, script, CI job, or backend worker to Revdoku over the REST API. Revdoku requires authentication before accessing buckets or storing files. AI clients should normally use hosted MCP with OAuth, and local coding agents should use the Revdoku CLI device flow. For every endpoint, see the complete Revdoku API documentation.

For a complete first request, use the JavaScript, Python, or C# (.NET) guide.

1. Create an account

Create an account at app.revdoku.com/users/sign_up, then sign in. Named API keys require a paid plan. Skill and MCP connections are available on Free within its limits. Revdoku does not create accounts through an agent or API call.

2. Choose the right authentication path

  • Hosted MCP: connect https://app.revdoku.com/mcp with Streamable HTTP and Revdoku OAuth. No API key is copied into chat.
  • Local agent or CLI: run revdoku login and approve the browser device flow. The CLI stores its credential locally.
  • Custom REST integration: create a reusable API key under Account → Access → Add connection → Create API key. To create new buckets, choose bucket_admin permission and access to all buckets in the intended account. Store the key in a secret manager or private environment variable.

Clients can discover supported agent authentication methods at GET /api/v1/agent_auth/capabilities. The preferred local flow is OAuth device authorization, which keeps account credentials and verification codes out of AI chat.

3. Authenticate REST requests

Point requests at the Revdoku base URL and send the API key as a bearer token:

export REVDOKU_URL=https://app.revdoku.com
export REVDOKU_API_KEY=revdoku_... # Get a key at https://app.revdoku.com/account/access
Authorization: Bearer $REVDOKU_API_KEY
Content-Type: application/json
Accept: application/json

4. Check the connection status

Confirm the key works and the connection is live with the status endpoint:

curl -fsS "$REVDOKU_URL/api/v1/status" \
  -H "Authorization: Bearer $REVDOKU_API_KEY"

A healthy connection returns your connection payload plus server_version and client_version; compare client_version against your installed connector to detect updates. Agents can call GET /api/v1/agent_auth/status, which returns the same connection payload. A missing or invalid key returns an authentication error with an error.code; reconnect through OAuth or create a replacement key in Account → Access.

Verify the connection from ChatGPT

If you are connecting an AI agent, use ChatGPT with Revdoku as a quick live check: ask it to save a test file in a bucket, read it back, and inspect its version history before building deeper API automation.

More details

This page covers connecting and verifying the connection. For every endpoint, creating buckets, uploading files, versions, access logs, incoming email, error codes, and hosted MCP, see the complete Revdoku API reference.

Markdown version
↧
Loading PDF…