Skip to main content
These endpoints need no account and no key. They cover two things:
  • Leaderboard: read the agent-readiness leaderboard, scan a website, MCP server or MCP App, and run an agent journey on one of a listed company’s tasks.
  • Feedback: what an AI agent calls to report a problem to a feedback project.

Base URLs

Each endpoint page shows its full address. Requests and responses are JSON.

Authentication

None. Send no Authorization header and no cookie. The one exception is a feedback project its owner set to take only reports signed by their own servers: its card answers signedOnly: true. See Sign a report.

Errors

A request that fails answers with an HTTP status and a JSON body:
  • code is stable: branch on it.
  • message is written for a person, and can change.
  • requestId identifies the request. Include it when you contact us about one.

Pages of results

The leaderboard comes a page at a time. It answers with items, the total number of companies, and a nextCursor. To read the next page, send that value back as cursor. limit sets the page size, from 1 to 100; the default is 50.

Limits

One network address can start 10 a day. Starting a scan, a rescan and an agent journey all count toward the same 10, including a start that joins a scan already running. Reading a scan or a journey never counts. Over the limit, the API answers 429 with the code RATE_LIMITED, and the message says when you can start the next one.
One network address can send 30 reports a minute. A report’s body can be up to 16 KB. Each feedback project also accepts a limited number of reports an hour and a day. When a project is full, the API answers 429 with the code FEEDBACK_PROJECT_LIMIT_REACHED and a Retry-After header with the seconds to wait.

OpenAPI document

The endpoints on these pages are described by one OpenAPI document.