ClawAudit verdict
agent-lightning
45
🟠 Risky
Significant concerns — only install if you understand the risks
The skill requires package installation and uses reinforcement learning, automatic prompt optimization, and supervised fine-tuning, which may pose risks if not used carefully.
⚠ 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.
93
security
90
transparency
90
maintenance
Permission integrity
Installs packages at runtime — transitive dependencies are not auditable
package_install
Findings (1)
Pattern match low
References agent configuration files
examples/train_agent.py · prose · downgraded · agent_config
Permissions & capabilities
No declared permissions — minimal attack surface.
package_install Is this flag fair?
Thanks — recorded.