os-improvement-loop
Warn
Audited by Socket on Apr 3, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The skill is mostly coherent with its stated purpose as a local multi-agent improvement workflow, and I found no clear credential theft or external exfiltration. However, it has a broad autonomous local footprint, can rewrite its own instructions, and explicitly instructs transitive installation of another plugin/skill, which raises trust and integrity concerns beyond a purely descriptive documentation skill.
Confidence: 79%Severity: 56%
Audit Metadata