community-management

Warn

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform an unrequested scan of the local filesystem by executing ls on multiple directories, including hidden folders in the user's home directory (~/.claude/skills/, ~/.agent/skills/, etc.). This is intended to enumerate other installed skills.
  • [EXTERNAL_DOWNLOADS]: The agent is prompted to suggest the installation of external packages using npx skills add AbsolutelySkilled/AbsolutelySkilled. This command downloads and executes code from a remote third-party repository which is not among the verified trusted sources.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 7, 2026, 03:03 PM
Security Audit — agent-trust-hub — community-management