Webmatik MCP server: run audits from Claude, Cursor, or any AI agent

We just shipped an audit check for AI-agent readiness — so naturally, we made Webmatik itself speak agent fluently. Two things landed today:

The Webmatik MCP server

Webmatik now runs a Model Context Protocol server at https://webmatik.ai/mcp. Connect it to Claude, Cursor, or any MCP-compatible agent and your assistant can:

• run_audit — start a full growth audit of any URL (SEO, performance, AI readiness, conversion, retention, UI/UX, accessibility, security),
• get_audit — fetch the scored result with the prioritized action plan once it's ready.

Auth uses your existing API key (generate one in your account, Starter or Growth plan) — pass it as a Bearer header or tool argument. Same engine, same quotas as the REST API. Discovery works out of the box: agents find the server via /.well-known/mcp.json.

webmatik.ai is now agent-ready

We also applied our own medicine after integrating the Agent Readiness checks:

• Markdown for agents — request any page with Accept: text/markdown and you get clean markdown instead of HTML (we coded the content negotiation ourselves — no paid add-ons).
• Content Signals in robots.txt — AI may search, cite, and learn from our content, now machine-readable.
• API catalog + Link headers (RFC 9727 / 8288) — agents can discover our API automatically.
• Agent skills index — a machine-readable how-to at /.well-known/agent-skills/index.json teaching agents to run Webmatik audits.

If you want your own site to pass these checks, run an audit — the AI Readiness tab now shows exactly where you stand.