Score your agent across 7 critical dimensions. Get a detailed scorecard in seconds. Fix everything with a single patch bundle.
Every agent has a unique fingerprint of strengths and weaknesses. We measure what matters.
Identity consistency, personality definition, contradiction detection across your SOUL.md
Skill organization, redundancy analysis, error handling coverage, boundary clarity
Prompt verbosity vs value ratio, context window usage, cost efficiency optimization
Prompt injection vectors, data exposure risks, input validation, auth boundary gaps
Task delegation clarity, agent boundaries, fallback chains, handoff protocol quality
Context persistence, retrieval relevance, state management, memory layer design
User-facing role clarity, voice consistency, purpose communication, boundary setting
Drop your SOUL.md, AGENTS.md, and skills list. No signup, no API key needed.
AI analyzes every dimension and returns a detailed scorecard with specific findings.
Pay 5 USDC on Base via x402. Receive a declarative patch bundle that fixes everything.
Paste your agent configuration below. Results in ~15 seconds.
Three endpoints. JSON in, JSON out. x402 for payments.
curl -X POST https://agentclinic.polsia.app/api/v1/audit \
-H "Content-Type: application/json" \
-d '{
"soul_md": "# MyAgent\nYou are a ...",
"agents_md": "# Routing\n...",
"skills_list": ["web-search", "code-review"]
}'