ClawAudit verdict
autoresearch-loop
as-autoresearch-loop
Describes a structured iterative improvement methodology for artifacts (skills, workflows, documents) — pure guidance framework with no execution of external code or data exfiltration.
⚠ 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 (3)
Uses eval() — can execute arbitrary code
SKILL.md · prose · downgraded · eval (
Instructs covert action — may act without user awareness
autoresearch-loop/eval_round7.md · prose · downgraded · silently
Popular HTTP library — network access
autoresearch-loop/eval_round2.md · prose · downgraded · got
Permissions & capabilities
No declared permissions — minimal attack surface.
Is this flag fair?
Thanks — recorded.