Skip to content
Back to search

MCP-VulnLab

Automated static AST + dynamic harness vulnerability fuzzer for open-source MCP servers, LangGraph tools, and LLM plugins, plus custom NVIDIA garak red-teaming probes.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/cuhawk--mcp-vulnlab/badge.svg)](/listings/cuhawk--mcp-vulnlab)

HTML

<a href="/listings/cuhawk--mcp-vulnlab"><img src="/api/listings/cuhawk--mcp-vulnlab/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 38 lines
  1. # MCP-VulnLab
  2. MCP-VulnLab reviews local MCP tool source and metadata and validates in-memory tool mocks.
  3. Install it with `pip install -e '.[test]'` on Python 3.12 or newer.
  4. ## Static source review
  5. Run `mcp-vulnlab . --sarif` to emit SARIF 2.1.0 or omit the flag for JSON advisories.
  6. The existing `--format sarif` option is also supported. Findings are review
  7. candidates. Python tracking is intraprocedural and conservative. The JS/TS
  8. pass matches source patterns in MCP modules without resolving data flow.
  9. The source auditor does not import or execute modules under review.
  10. ## Offline mock validation
  11. Use `await run_mock(tool, arguments, schema, path_fields=("path",))` from
  12. `mcp_vulnlab.schema_validator` with a trusted async mock that takes one
  13. argument dictionary. It applies a compact UTF-8 JSON byte budget, validates
  14. against JSON Schema Draft 2020-12, and normalizes selected top-level path
  15. fields within a virtual POSIX root. The timeout covers the callable and
  16. requires cooperative cancellation. Schema references resolve within the
  17. document. Path normalization performs no filesystem access.
  18. ## Tool metadata review
  19. Run `python -m mcp_vulnlab.garak_probes.mcp_tool_shadowing tools.json --baseline trusted.json`.
  20. Each file contains a JSON array of tool records with a `name` and optional
  21. `description`. The baseline is optional. The scorer flags instruction-override
  22. patterns, duplicate normalized names, and descriptions that differ from the
  23. trusted baseline. A score of 1 indicates a review candidate.
  24. To use the optional adapter with a garak installation that provides
  25. `always.Passthru`, place `mcp_tool_shadowing.py` in that installation's
  26. `garak/probes` directory and install MCP-VulnLab in the same Python
  27. environment. Select `mcp_tool_shadowing.MCPToolShadowing` and configure
  28. its `tools` and optional `baseline` lists under
  29. `plugins.probes.mcp_tool_shadowing.MCPToolShadowing`. The probe scores the
  30. configured metadata and records the results without calling the generator.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.