quality-retrospective
Warn
Audited by Socket on Apr 17, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The skill is largely coherent for retrospective analysis and uses only local project data, with no obvious credential theft or external exfiltration. The main risks are transitive trust in an unverified helper skill, broad artifact ingestion that could carry prompt-injection content, and optional unattended routing/writes via `--auto-yes`.
Confidence: 82%Severity: 56%
Audit Metadata