implement-plan
Warn
Audited by Socket on Aug 21, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the core purpose is legitimate, but the skill expands execution through default delegation to subagents/plugins/other skills and optionally mixes external research with write/exec capabilities. No direct malware, credential theft, or installer payload is present, but the transitive-trust and prompt-injection surfaces make it higher risk than a local-only implementation guide.
Confidence: 82%Severity: 57%
Audit Metadata