kb-librarian
Warn
Audited by Socket on Aug 15, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill’s file access and knowledge-base maintenance behavior mostly fits its stated purpose, but it depends on an unverified local script (~/.claude/scripts/kb-index.py) and ingests arbitrary external content while retaining write access. There is no clear evidence of credential theft or malicious exfiltration, but the unverifiable executable and prompt-injection exposure make the skill high risk.
Confidence: 86%Severity: 78%
Audit Metadata