ClawAudit verdict
skill-mcp-security-audit
mcp-security-audit
Fetches from network AND uses eval/Function
Security audit skill teaching defensive MCP practices; commands shown are grep and npm audit patterns for auditing other software and the skill itself does not exfiltrate data.
Automated static analysis — not a human review. ClawAudit flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.
What it does
These are capability combinations: each listed behavior occurs in the skill, but ClawAudit detects co-occurrence — it does not verify that one flows into another. Read the code to confirm a live chain.
Fetches from network AND uses eval/Function — the remote-code-execution pattern (data-flow not verified)
LLM05 · LLM06 · ASI05
Receives external input AND uses eval — the remote code-injection pattern (data-flow not verified)
LLM01 · LLM05 · ASI01 · ASI05
Receives external input AND executes processes — the shape of a command & control channel
LLM05 · LLM06 · ASI10
Accesses system credential store AND makes external network calls
LLM02 · ASI03
Executes processes AND makes external network calls — may exfiltrate command output
LLM02 · LLM06 · ASI03
Accesses credentials AND makes external network calls — potential credential theft
LLM02 · ASI03
Permission integrity
network_out
credential_access
Findings (8)
Uses eval() — can execute arbitrary code
README.md · code · eval(
Possible hardcoded credential
README.md · code · token=' + process.env.API_KEY)
// Unrestricted file access — reads any file on
Base64 decode (atob) — may hide malicious payloads
README.md · code · atob(
Uses exec() — may execute shell commands
README.md · code · exec(
Dynamic import() — loads module at runtime
README.md · code · import('
Popular HTTP library — network access
SKILL.md · code · axios
Accesses sensitive environment variables
README.md · code · process.env.API_KEY
Makes HTTP request to external URL
README.md · code · fetch('https://
Why the tier is capped
Execution sink present in raw bytes (Hard Floor: class D). Final tier capped at Caution — cannot be lifted by any downgrade, example-payload opt-in, or allowlist.
Permissions & capabilities
No declared permissions — minimal attack surface.
network_incredential_storenetwork_outprocess_execdynamic_evalcredential_access Thanks — recorded.