collection-claude-code-source-code
Warn
Audited by Socket on May 17, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill's educational purpose partly matches the content, but its actual footprint goes further by instructing installation and execution of code from an unverified personal GitHub repository described as leaked/decompiled Claude Code. Official API usage in examples is normal, yet combining untrusted code installation with environment-key usage and agent command execution makes the overall risk medium-high.
Confidence: 88%Severity: 78%
Audit Metadata