ClawAudit verdict

self-improving-agent

kay-self-improving

45
🟠 Risky
Significant concerns — only install if you understand the risks

A self-modifying agent with hooks that auto-trigger on all skill events, reads/writes skill files, and can submit PRs; the broad auto-modification scope and hooks into every skill lifecycle represent unusual elevated capability beyond a normal skill's purpose.

⚠ 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.

80
security
80
transparency
70
maintenance

Findings (1)

Pattern match high

Pipe to bash — executes piped content as shell commands

SKILL.md · prose · downgraded · | Bash

Why the tier is capped

Execution sink present in raw bytes (Hard Floor: class A). 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.

network_in

Is this flag fair?

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