What it does
This MCP server gives AI coding assistants such as Claude Desktop, Cursor, Windsurf and Cline two audit tools. audit_mcp_repository runs a static AST-based audit on a public Git repository containing an MCP server, checking for command injection (CWE-78), path traversal (CWE-22), hardcoded secrets (CWE-798), missing authentication on HTTP/SSE transports (CWE-306) and tool and prompt injection (CWE-1336). audit_local_directory audits a local directory of MCP server source code.
It is intended for checking an MCP server or tool before installing or running it. Audit results include a trust score from 0 to 100, a letter grade, line-by-line findings with CWE identifiers and remediation instructions, and a markdown report. The local directory audit returns a structured JSON record and a remediation checklist. The publisher says the analysis covers Python, TypeScript and JSON-RPC implementations.
Setup means adding the server to the client's MCP configuration, for example as a command entry that runs python3 against server.py, or running it in Docker with the provided Dockerfile. The publisher states that the analysis is static and that the audited code is never executed.
Limitations and requirements
The repository audit requires a public Git URL. The local directory audit requires a filesystem path to the target directory. The command-based setup requires python3 and an absolute path to server.py.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Official MCP Registry first seen 13 Nov 2026 · last seen 13 Nov 2026 · its page on the source
MCP Security & Vulnerability Auditor
Static AST security scanner detecting command injection, leaked secrets, and SSRF in MCP tools.
v1.0.0As shown on registry.modelcontextprotocol.io when we last read it