Skillv1.0.0

ClawScan security

Secret Detector · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignApr 30, 2026, 1:49 AM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's instructions and requirements are consistent with a local secrets-scanning helper; it doesn't request credentials or install code, but its provided commands are narrow and may miss some secrets.
Guidance: This skill appears coherent and low-risk: it only provides local grep/Select-String commands to find secrets and requests no credentials. Before using it, note that the example commands are narrow (only look for GitHub/AWS tokens in .js/.py files) and will miss other patterns listed in the file (private keys, Slack tokens, generic API keys, URLs with passwords). If you plan to use it, either extend the grep patterns and file globs to cover other file types (e.g., .env, .yaml, .pem, config files) or use a dedicated vetted tool (gitleaks, git-secrets, truffleHog). If the scan finds secrets, treat them as compromised: rotate keys immediately, remove them from the repo history, and avoid pasting secrets into chat or external services.

Purpose & Capability: okName/description (scan for leaked secrets) match the content of SKILL.md. The patterns listed are appropriate for secret detection. The skill is instruction-only and does not attempt to access unrelated services or credentials.
Instruction Scope: noteInstructions contain useful regex patterns and example commands to run locally (PowerShell Select-String and grep). However the provided commands only search for two token types (GitHub and AWS) in .js/.py files, while the 'Patterns to Detect' section lists many more patterns (private keys, slack tokens, generic API keys, passwords in URLs). This mismatch means following the commands as-is could miss many secrets. The instructions do not instruct reading unrelated system files or sending results externally.
Install Mechanism: okNo install spec and no code files — instruction-only, so nothing is downloaded or written to disk by the skill itself.
Credentials: okNo environment variables, credentials, or config paths are requested. That is appropriate for a local scanning helper.
Persistence & Privilege: okalways is false and the skill does not request persistent system presence or modify other skills. Autonomous invocation is allowed (platform default) but not combined with other risky behaviors.