mcp-audit

Static security + correctness audit for MCP server repos, in Python and TypeScript/JavaScript.

pip install mcp-audit
mcp-audit                     # scan the current directory
mcp-audit /path/to/repo       # scan a specific repo
mcp-audit --json              # machine-readable output
mcp-audit --check fastmcp_wrapper_layer  # one check only
mcp-audit --list-checks       # list available checks

Exit codes: 0 clean, 1 at least one finding, 2 usage error, 3 nothing examined.

"0 findings" is not the same as "0 files"

A scanner that cannot tell those apart is worse than no scanner, because it issues a clean bill of health for code it never opened. mcp-audit has shipped that bug twice, so it now reports how many files it read and exits 3 when that number is zero:

$ mcp-audit ./docs
mcp-audit: examined 0 files under ./docs - nothing here to audit. This is NOT
a clean result: no Python or TypeScript source and no dependency manifest was
found. Check the path, or pass --allow-empty if that is expected.

$ mcp-audit ./server
mcp-audit: OK - 0 findings across 34 file(s) under ./server

Pass --allow-empty to exit 0 instead, where a repo legitimately holds no auditable code. In the GitHub Action, exit 3 fails the job even in report-only mode: a green check that read nothing is the worst result the action can produce.

Related: build directories (dist, build, out) are skipped as generated residue, unless nothing else is there. A published npm package ships package/dist/*.js and no sources at all, so pointing mcp-audit at an extracted tarball scans the compiled output and says so. Because the fallback is reachable only from an otherwise-empty scan, it can never add noise to a repo that has real sources.

What it checks

| ts_dns_rebinding | HIGH | TypeScript. StreamableHTTPServerTransport / SSEServerTransport constructed without DNS-rebinding protection, in a repo with no authentication check. enableDnsRebindingProtection defaults to false on the constructor (GHSA-w48q-cv73-mx4w / CVE-2025-66414, CVSS 7.6), so the transport accepts any Host and any Origin. A page the victim visits can re-point its own hostname at 127.0.0.1 and call the endpoint as same-origin, reaching tools/list and tools/call at local privilege. Binding to localhost does not mitigate it, and CORS never runs. Suppressed by enableDnsRebindingProtection: true + allowedHosts/allowedOrigins, by an independent host gate or hostHeaderValidation(), by any repo-wide auth check, and entirely for stdio-only servers. Added in v0.9. | | ts_command_injection | HIGH | TypeScript. An MCP tool parameter reaching child_process.exec/execSync, or spliced into a shell command string, or landing in the executable position. The safe array-argv-without-shell: true pattern is correctly not flagged, and a value is only counted when the splice is visible (a template hole, a + concat, or the raw parameter) — an allow-list table lookup does not fire. Added in v0.9. | | ts_destructive_fs_sink | MEDIUM | TypeScript. An MCP tool parameter reaching fs.rm/rmSync/unlink/rmdir/promises.rm or rimraf with no path-containment guard. Suppressed by canonicalize-and-confine (path.resolve + startsWith/relative), by an allow-set, by a closed-set schema (z.enum/literal/number) that cannot express traversal, and when the path came back from an opaque call rather than from the caller. Added in v0.9. | | ts_tool_input_validation| LOW | TypeScript. MCP tool input schemas that constrain nothing: z.any(), a bare z.string() with no length/pattern/format, a handler parameter typed any, or a registerTool whose handler destructures arguments it never declared. Recognizes constraint methods by exclusion — only provably transparent wrappers (.describe(), .optional(), .default(), …) count as unconstrained — so new zod formats degrade to false negatives, not false positives. Hygiene, not a vulnerability: expect findings on well-written servers. Added in v0.9. |

Two engines, one scan

The six checks above the divider read Python via ast; the four ts_-prefixed checks read TypeScript/JavaScript via a stdlib-only masking lexer (mcp_audit.jsparse). Both engines run on every scan — a repo may contain both, and a scanner that silently skipped one language would hand back a clean bill of health it had not earned. Before v0.9 that is exactly what mcp-audit did to every TypeScript MCP server: OK — 0 findings, without opening a file.

There is no parser dependency and never will be one: pip install mcpdone-audit must not pull node, tree-sitter, or a wasm blob. The lexer instead produces a masked copy of the source, byte-for-byte the same length, with comments and string contents blanked so offsets and line numbers stay exact. Template-literal ${...} holes stay live, since exec(`git checkout ${branch}`) is the shape the injection check exists to find. When the lexer cannot walk a file cleanly it marks it degraded and every check returns nothing for it — a mis-lexed file must never produce a finding.

More checks are landing — hard-coded secrets leaked into tool output, write-API tools missing a FORBIDDEN_NAMES-style guardrail, path traversal in filesystem-touching servers, and SSRF via tool-controlled URLs.

Output format

$ mcp-audit examples/bad/
[HIGH  ] starlette_badhost @ uv.lock
           uv lockfile pins starlette==0.36.3 — vulnerable to BadHost (CVE-2026-48710). Patched in 1.0.1.
           -> Upgrade Starlette to >=1.0.1 (the BadHost patch). If FastAPI pulls Starlette transitively, pin it explicitly. ...

[HIGH  ] fastmcp_wrapper_layer @ server.py:18
           tool 'fetch_url' (def) calls asyncio.run() inside its body. FastMCP invokes tools inside an already-running event loop, and asyncio.run() raises RuntimeError when nested. This will fail at the first real MCP protocol call even if every unit test passes.
           -> Convert the tool to `async def` and replace `asyncio.run(...)` with `await`. ...

mcp-audit: 2 finding(s) — 2 high across 11 file(s)

--json emits one object: {"root": "...", "finding_count": N, "files_examined": N, "scanned_build_output": false, "findings": [...]}. Each finding has check, severity, path, line, message, remediation.

files_examined is there so a consumer can assert the scan actually read something rather than trusting finding_count: 0.

Use in CI (GitHub Action)

Drop MCP security scanning into any repo's CI. The job fails on findings (exit 1); flip fail-on-findings for a report-only gate.

# .github/workflows/mcp-audit.yml
name: mcp-audit
on: [push, pull_request]
jobs:
  audit:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: Alienbushman/mcpdone-samples/mcp-audit@mcp-audit-v0.9
        with:
          path: .                 # dir to scan (default '.')
          # version: 0.8.0        # pin the mcpdone-audit release (default: latest)
          # package: ./mcp-audit  # any pip spec: local path, VCS ref, private mirror
          # args: --check command_injection
          # fail-on-findings: false   # report-only: annotate but keep the job green

The action exposes a finding-count output for downstream steps. It wraps the same mcpdone-audit PyPI package, so local runs and CI runs are identical. Prefer pinning version: in CI for reproducibility.

A scan that examines 0 files fails the job (exit 3) even with fail-on-findings: false, because a green check that read nothing is worse than a red one. Pass --allow-empty via args where that is expected.

package: overrides version: and takes any pip install spec. It exists because the action installs from PyPI by default, which means nothing can exercise a fix that has not been released yet. This repo's own self-test hit exactly that: three new assertions failed against the published 0.9.0 while the repo was already fixed. Same shape as the bug above, one layer out.

You can also run it via pre-commit or as a plain pip install mcpdone-audit && mcp-audit step in any pipeline.

What this is not

• It is not a runtime sandbox. Static analysis only.
• It does not install your venv to introspect it. It reads what's declared (manifests + lockfiles + source).
• It will not detect every vulnerability — only the classes its checks know about. Treat zero findings as "no known issues from this tool," not as a clean bill.

Background

• BadHost write-up: https://mcpdone.com/blog/badhost-mcp-servers
• FastMCP wrapper-layer bug write-up: https://mcpdone.com/blog/fastmcp-wrapper-layer-bug

Development

git clone https://github.com/Alienbushman/mcpdone-samples
cd mcpdone-samples/mcp-audit
pip install -e ".[dev]"
pytest
python smoke_test.py

To add a check: drop src/mcp_audit/checks/<name>.py exposing a module-level CHECK_ID and a check(root: Path) -> list[Finding] callable. Register it in src/mcp_audit/checks/__init__.py. Add fixtures + tests under tests/.

License

MIT.