Guide

AI Readiness Audit Checklist for Websites and APIs

Most teams want AI traffic but are missing basic machine-readable surfaces. This checklist helps you ship the minimum public signals AI agents need to discover your product, understand your docs, and call your API safely.

1. Discovery surfaces

Publish /.well-known/openapi.json and keep it reachable with a 200 response.
Add /.well-known/air.json with contact, legal, and verification metadata.
Publish /llms.txt with canonical docs, API base URL, and endpoint references.
Avoid blocked robots rules for docs and machine entrypoint files.

2. Callable API surfaces

Use valid OpenAPI schema with request and response examples.
Keep endpoint paths stable and return deterministic JSON shapes.
Expose clear error payloads with status code guidance.
Avoid auth-only landing pages for your public docs.

3. Documentation for LLM ingestion

Keep a canonical docs index page with stable URLs.
Link directly from docs to API reference and quickstart.
Provide short, copyable cURL examples for core endpoints.
Keep markdown assets crawlable and lightweight.

4. Trust and safety signals

Publish privacy policy, terms, and cookie policy on stable URLs.
Expose contact email and support route in public metadata.
Keep attestation fields aligned across air.json and docs.

5. Reliability checks

Re-test core endpoints from multiple origins to detect flaky behavior.
Prevent redirect loops and temporary anti-bot blocks on docs pages.
Keep timeouts and response sizes predictable.

Fast implementation plan

Day 1: publish OpenAPI, air.json, and llms.txt with canonical links.

Day 2: harden docs with quickstarts and schema-backed examples.

Day 3: run a public audit and fix the top failing checks by severity.

Run a live AI readiness audit on Agentability →