ClawAudit verdict

self-refine-reflection

88
🟢 Trusted
Low risk — reviewed by ClawAudit, behavior matches stated purpose

The skill is designed for systematic self-reflection and iterative output refinement. It does not contain any apparent malicious code or intent.

⚠ Flagged for review — coarse, uncorroborated signal, not a confirmed exploit. Review the config yourself before installing.

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.

58
security
70
transparency
70
maintenance

Findings (5)

Pattern match medium

Instructs covert action — may act without user awareness

SKILL.md · prose · downgraded · silently

Pattern match medium

References sudo — requests elevated privileges

BENCHMARK.md · prose · downgraded · sudo

Pattern match medium

HTTP request to bare IP address — common in malicious payloads

BENCHMARK.md · prose · downgraded · http://127.0.0.1

Pattern match medium

References agent configuration files

README.md · code · CLAUDE.md

Pattern match low

Opens WebSocket connection

BENCHMARK.md · prose · downgraded · WebSocket

Permissions & capabilities

No declared permissions — minimal attack surface.

Is this flag fair?

Check another skill Browse the registry Auditing your own skills or configs? Use the API