Five working walkthroughs, every one against the live API. Copy-pasteable curl, Python and TypeScript; the real endpoints, headers and tier limits; and no code that calls something we do not actually ship. The free tier covers everything here except webhooks.
Connect the AI Law Tracker MCP server to Claude Code and Cursor in about two minutes. Real config JSON, the actual tool names, the free tier, and exactly how your API key travels.
Streamable HTTP · https://ai-law-tracker.com/mcpRead the guide →Pull every tracked AI law for a US state from a free JSON API — with working curl, Python and TypeScript, real pagination, the tier caps that apply, and the error bodies you must handle.
GET /v1/laws · GET /v1/states · GET /v1/searchRead the guide →Subscribe to law.created and law.updated, verify the HMAC-SHA256 signature exactly the way the sender computes it, and de-duplicate at-least-once deliveries. Real headers, real algorithm, real retry policy.
POST /v1/webhooks · X-ALT-Signature · sha256Read the guide →Turn AI-law effective dates into a live calendar your team can subscribe to. Free JSON and RFC 5545 iCal feeds, a cron alerting example, and why raw_deadline and date_precision matter more than the parsed date.
GET /v1/deadlines · ?format=ical · free tierRead the guide →Wire /v1/search and /v1/laws into a RAG or agent loop so every legal claim is retrieved, not remembered — with official_url citations, verifiable sources, and a refusal path when retrieval comes back empty.
GET /v1/search · /v1/laws · official_url citationsRead the guide →No card required. A free key gives you 60 requests/minute, 2,000 requests/day and the complete record envelope. Anonymous access works too, at lower limits — you can run the first example before you sign up for anything.
Data by AI Law Tracker, licensed CC BY 4.0 — attribution required: “Data by AI Law Tracker (CC BY 4.0) — https://ai-law-tracker.com”. Informational data, not legal advice.