Agent Identity Trust MCP Server

This MCP server contains multiple critical security vulnerabilities that substantially undermine trust and user safety. The most severe issue is an unsafe filesystem path traversal in the sys.path manipulation at module load, combined with reliance on an external shared authentication middleware that cannot be verified. Additionally, the server makes unauthenticated network calls to external metering endpoints (proofof.ai) without user consent or clear disclosure, creating a covert data exfiltration risk. The in-memory credential store is not persistent and offers no real security guarantees. While the code quality is moderate and the stated purpose (agent identity/DID management) is legitimate, the authentication bypass patterns, external phone-home calls, and unsafe path manipulation create an environment where this server could easily be abused or compromised. Supply chain analysis found 3 known vulnerabilities in dependencies (0 critical, 3 high severity). Package verification found 1 issue.

7 files analyzed · 16 issues found

Security scores are indicators to help you make informed decisions, not guarantees. Always review permissions before connecting any MCP server.