session-bootstrap
Warn
Audited by Socket on Sep 21, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The stated purpose matches the read-only status-check behavior, and the flagged command substitution is benign documentation context. However, the skill executes an unverifiable custom local script (`~/.claude/evolution/bin/evolve-status.sh`) tied to an unverified local framework not found in official Anthropic docs, so install/execution trust is weak even without overt exfiltration or credential handling.
Confidence: 90%Severity: 72%
Audit Metadata