optimizer
Warn
Audited by Socket on Mar 28, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill's core purpose broadly matches code analysis and patching, and the visible credential/data flows are mostly proportionate and official. However, it grants an agent powerful self-modification behavior, mandatory dynamic registration, terminal-enabled execution, and reads other skills' instructions while retaining write access, creating meaningful autonomy and indirect prompt-injection risk. Supply-chain risk is moderate due to official but unpinned Playwright MCP installation.
Confidence: 84%Severity: 68%
Audit Metadata