jaimenbell-mcp-security-scanner

# mcp-security-scanner [](https://github.com/jaimenbell/mcp-security-scanner/actions/workflows/ci.yml) A static security scanner for [Model Context Protocol](https://modelcontextprotocol.io) servers. Point it at an MCP server repo; it reads the source and flags the vulnerability classes that actually show up in production MCP servers — with a severity, a `file:line`, a remediation, and an honest **confidence** on every finding. > [!info] What this is, plainly > This is **static analysis**, not a prover. It reads code; it does not run your server, and it does not prove any finding is remotely exploitable. It produces a prioritized review queue, not a verdict. A "clean bill" means *these detectors found no critical/high patterns* — not a security guarantee. That boundary is printed on every report on purpose. > [!warning] Not to be confused with the other PyPI package named `mcp-security-scanner` > There is an unrelated project on PyPI under the plain name `mcp-security-scanner` (a runtime pentester that connects to a *live* MCP server over HTTP/SSE). This repo is a different tool: it performs **static analysis of server source code**, offline, with no network connection to the target. Because the plain name was already taken, this project's PyPI distribution is published as `jaimenbell-mcp-security-scanner`; the console command (`mcp-scan`) and the import package (`mcp_scanner`) are unaffected. ## What it scans Seven detector families. The first six are grounded in a real finding from a fleet-wide audit of production MCP servers; the seventh (added 2026-07-21) covers scheduled jobs, wrappers, and IaC/CI files — cron, systemd, GitHub Actions, PowerShell/bash/batch deploy scripts: | # | Class | Detects | |---|---|---| | 1 | **Codegen / template injection** | Jinja `autoescape` off in a code-*generating* tool that renders untrusted fields into generated …