SkillAudit report — Azure/azure-mcp

Scanned 2026-04-24 by SkillAudit v0.2 (static checks + LLM-assisted prompt-injection red-team).

Commit: 54f04ab · Stars: 1213 · Days since last push: 76

LLM prompt-injection probe: no-tool-surface

Overall grade: F (40/100)

Security findings

Production sources:

• HIGH eng/npm/wrapper/index.js:46 — child_process.exec with template-string interpolation — argv is concat-built, not escaped

execSync(\npm install ${platformPackageName}@latest\, {

• HIGH eng/npm/wrapper/index.js:55 — child_process.exec with template-string interpolation — argv is concat-built, not escaped

execSync(\npm install ${platformPackageName}@latest --no-save --prefer-online\, {

Permissions

_No findings on this axis._

Credentials

Production sources:

• WARN eng/tools/ToolDescriptionEvaluator/.env.example — .env file present in repo tree — verify it's a template, not real secrets

eng/tools/ToolDescriptionEvaluator/.env.example

Maintenance

Production sources:

• HIGH (meta) — Repository is archived — no fixes will be issued upstream

archived

Compatibility

Production sources:

• WARN (meta) — No engines (Node) or python_requires declared — cross-client compatibility unverified

Documentation

Production sources:

• WARN README — README has no install section

no install section

• WARN README — README has no usage/example section

no usage section

• WARN README — README has no fenced code blocks — no copy-pasteable command

no code blocks

Methodology

SkillAudit v0.2 clones the repo at the provided ref (default: default branch, HEAD) into an ephemeral sandbox, runs six static checks over .js/.ts/.py sources, queries the GitHub API for maintenance signals, and runs an LLM-assisted prompt-injection red-team over the MCP tool surface. Each axis is scored against the rubric at .

The prompt-injection axis extracts each server.tool(...) / @app.tool registration + the first ~60 lines of handler body, hands them to Claude Haiku 4.5 with a red-team system prompt, and asks for structured findings on untrusted-content flow into tool responses. One API call per scan, bounded at ~15K input tokens.

How to improve this grade

• Security — static: validate tool-input URLs against an allowlist before fetch/axios calls; use execFile with argv arrays instead of exec with template strings; never pass untrusted strings to subprocess with shell=True.
• Security — prompt injection: never return fetched web-page / file / email content verbatim in a tool response. Wrap with a framing marker (e.g., <untrusted-content>...</untrusted-content>), summarize rather than inline, and never let untrusted content share a turn with credentials or other tool output.
• Credentials findings: redact env-var reads before log lines and error messages; treat any string that ends up in a tool response as public.
• Maintenance: if the repo is inactive, document the maintenance model — "MCP tool, no breaking changes expected" is a legitimate signal.
• Docs: add a README install + usage section with a copy-pasteable command; add a SECURITY.md with a disclosure channel.

_Report generated by skillaudit.dev_