Open-source MCP scanner combines static analysis and live prompt-injection tests
KookyTax5493 · reddit · 2026-07-29
An open-source tool, mcp-scanner, targets a growing security problem for MCP servers: prompt injection through tool outputs.
- It combines static analysis and live probing.
- Static checks look for shell execution, unsafe deserialization, hardcoded secrets, unscoped tools, and missing input validation.
- Live tests connect as a real MCP client and fire prompt-injection payloads such as instruction override, role hijack, exfiltration, tool-chaining abuse, encoding tricks, and homoglyph attacks.
- Responses are judged with a two-stage system: keyword pre-filtering plus an LLM judge fallback.
- The author says it successfully caught an issue in the official mcp-server-fetch, where a poisoned page was echoed back unsanitized.
The project is open source on GitHub and the author is asking for production feedback on what would make it useful in real deployments.
Related event: Open-Source MCP Scanner Detects Injections and Leaked Secrets(2 posts)→
More from coding & agent
- Connected steps are not a workflow: nobody designs the handoff — usehive · 2026-10-16
- Making Blender animations with Opus, Astra and Argon: how you prompt still matters — chongdashu · 2026-10-16
- Running PolyStrata + Cline coding agent on a $0-fancy laptop at ~90 tok/s — rpwoerk · 2026-10-16
- How do you attribute agent gains: model training or infrastructure? — jrxna · 2026-10-16
- Feed Claude your books, blogs and top engineers' code for claude.md — 'it now writes better code than me' — blaizedsouza · 2026-10-16
- Open-source /label plugin auto-names and colors Claude Code sessions in one step — hermanschutte · 2026-10-16