MCP answer-quality canary

Enables security testing of MCP-speaking agents by serving as a malicious MCP server that plants canaries and catches evidence of tool poisoning, exfiltration, and other hijack techniques, then reports resilience scores.

A continuous, out-of-band trust and reliability layer for the MCP ecosystem. It fingerprints MCP server tool definitions, detects and classifies drift (e.g., rug pulls) via a severity taxonomy, maintains a hash-chained evidence ledger, and gates CI with SARIF—while also acting as an MCP server itself so agents can check a server's safety before binding.

Lets agents check remote MCP servers before calling a tool, returning a safe/caution/unsafe verdict with live status, latency, auth requirements, spec readiness and whether the tool list just changed in a breaking way. It also suggests alternative servers for a given capability and sends webhook alerts when a watched server's verdict, auth state, or tool list changes.

A local-first, deterministic, read-only MCP server that audits test suites for false-green tests, tautological assertions, and mock-contract drift, ensuring tests truly validate production code. It provides tools to detect test fidelity issues, verify mock drift, and synthesize strict mock contracts.