We have spent years building reputation systems for builders: turning projects, contributions, and verified signals into something people can trust. Polygraph.so applies that same instinct to the agentic web.
The product is still early, and the grading system has a
AI agents need to know which tools they can trust before installing them.
We test third-party AI tools for hijacking, permission overreach, and data leaks.
Then we publish a public A to F grade with evidence attached.
Open-source work, supported by $POLYGRAPH fees from the
polygraph.so
polygraph.so
Behavioral polygraphs for AI agents and MCP servers — grades anyone can re-run. No graded party pays us.
