Caveat verdict

chaos-lab

45
🟠 Risky
Significant risk patterns flagged — automated deep scan, not behavioral proof.

The skill's use of credential_access and package_install capabilities raises concerns about potential security risks, especially in a multi-agent framework for exploring AI alignment.

⚠ 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.

55
security
70
transparency
90
maintenance

Permission integrity

Installs packages at runtime — transitive dependencies are not auditable

package_install

Findings (3)

Pattern match high

Possible hardcoded credential

scripts/run-duo.py · prose · downgraded · API_KEY="): API_KEY = line.strip().split(

Pattern match medium

pip3 install — installs Python packages at runtime

SKILL.md · code · pip3 install

Pattern match medium

Popular HTTP library — network access

examples/flash-results.md · code · got

Permissions & capabilities

No declared permissions — minimal attack surface.

credential_accesspackage_install

Is this flag fair?

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