coding-agent-hardened
Warn
Audited by Socket on Apr 21, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the skill is largely coherent as a coding-agent orchestration guide, but its footprint is high-risk because it empowers autonomous code changes and networked repo actions, and it processes untrusted PR/repo content with write/exec capabilities. Install provenance is mostly normal; the main concerns are autonomy and prompt-injection exposure rather than confirmed malware.
Confidence: 86%Severity: 62%
Audit Metadata