proactive-agent
Warn
Audited by Socket on Mar 28, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core memory and recovery behavior is coherent with the stated purpose, and most data flow is local. The main concerns are the encouragement of autonomous background actions, expansive 'use every tool' guidance, and medium supply-chain risk from an unpinned GitHub-based skills install path with limited provenance. No direct credential harvesting, exfiltration endpoint, or confirmed malicious behavior is evident.
Confidence: 83%Severity: 56%
Audit Metadata