Caveat verdict

self-improving-intent-security-agent

88
🟢 Trusted
No high-risk patterns surfaced by the deep scan — automated capability review, not behavioral proof.

Documentation-first toolkit providing templates and local helper scripts; explicitly states helpers operate on local files only.

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

Automated static analysis — not a human review. Caveat flags capabilities, not confirmed intent, and can produce false positives. Disagree with this verdict? Use Dispute below.

58
security
100
transparency
80
maintenance

Findings (3)

Pattern match critical

Possible hardcoded credential

references/deployment.md · code · TOKEN="your-token

Pattern match medium

References agent configuration files

CLAUDE.md · code · CLAUDE.md

Pattern match low

References webhook/callback URL

CLAUDE.md · prose · downgraded · WEBHOOK_URL

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class B/F). 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.

Is this flag fair?

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