// the free scan
Agent Readiness Scan
See what AI agents actually do with your tools.
Instant analysis of your tool surface. No API calls made.
How it works
Quick Scan runs rule-based static analysis across five dimensions — presentability risk, description clarity, input friction, parameter discoverability, and error recoverability — calibrated against thousands of real agent-tool interactions. No LLM calls.
Full Scan auto-generates natural-language prompts from your tool surface, loads all your tools into Claude Sonnet, and reports exactly which tools the model selected, which it ignored, and what parameters it sent — including any it hallucinated.
No API calls are made to your service. Nothing is stored. Source on GitHub.
Quick Scan runs rule-based static analysis across five dimensions — presentability risk, description clarity, input friction, parameter discoverability, and error recoverability — calibrated against thousands of real agent-tool interactions. No LLM calls.
Full Scan auto-generates natural-language prompts from your tool surface, loads all your tools into Claude Sonnet, and reports exactly which tools the model selected, which it ignored, and what parameters it sent — including any it hallucinated.
No API calls are made to your service. Nothing is stored. Source on GitHub.
Scanning tool surface...
We don't store your tool definitions or scan results. Full scans send your tool definitions to Anthropic's API for analysis — see Anthropic's privacy policy. This scan is for informational purposes only and doesn't constitute a comprehensive security or quality audit. Read our full privacy policy.