agentspeed.

Documentation

One workflow for site owners — scan, monitor, investigate, certify — plus the telemetry API for teams building agents.

For site owners — the readiness workflow
The scoring rubric
Every check, weight, and threshold behind the 0–100 score — the same published rubric at every stage: scan, monitor, investigate, certify.
Readiness badge
Embed your AI-agent readiness score on your own site.
MCP server
Point Claude Code, Claude Desktop, or Cursor at one URL and ask whether a site is readable by AI agents. No key, no signup.
For agent builders — telemetry
Report a run: ingest API
Send agent telemetry with one authenticated POST. Copy-paste Python, TypeScript, and curl examples. Metadata only — no prompt or output content, by design.
Documentation · AgentSpeed