Is your AI agent actually ready — or just a demo that talks well?
Paste a system prompt. Get a structural production score in seconds. Fix gaps in one click. Works for builders and buyers checking AI vendors.
⏱ Under 30 seconds🔒 Stays in your browser📤 Share a number people understand🛠 Fix without installing anything
Structural hygiene only — not a red-team or certification. Same engine as the free CLI for teams that want CI later.
I have an AI agent
I’m buying AI software
From ChatGPT custom instructions, Claude projects, Cursor rules, LangChain, CrewAI, internal agents — anything that tells the model who it is.
10 diligence questions. Self-attested — not a formal audit. Get a score + red flags you can paste into LinkedIn or a buying thread.
Want this in CI? pip install crewscore ·
uses: shmindmaster/crewscore@v1 Open source on GitHub
· structural scan only · templates must be paired with runtime gates
Same scoring rules in the browser and the CLI.