ClawAudit verdict
improvement-discriminator
The skill provides a clear description of its purpose and usage, and its capabilities are used legitimately for evaluating and scoring improvement candidates.
⚠ 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.
Findings (4)
Possible hardcoded credential
interfaces/critic_engine.py · prose · downgraded · password="DEMO_ONLY_NOT_FOR_PRODUCTION
Uses eval() — can execute arbitrary code
scripts/score.py · prose · downgraded · eval(
Uses exec() — may execute shell commands
scripts/score.py · prose · downgraded · exec(
Popular HTTP library — network access
interfaces/assertions.py · prose · downgraded · got
Permissions & capabilities
No declared permissions — minimal attack surface.
Is this flag fair?
Thanks — recorded.