aggressively-cleanup-skills

Pass

Audited by Gen Agent Trust Hub on Aug 23, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Node.js scripts (audit.mjs, report.mjs, archive.mjs) to perform its primary functions. These scripts interact with the local file system to read configuration files and move skill directories to an archive location.\n- [INDIRECT_PROMPT_INJECTION]: \n
  • Ingestion points: Session history files are read from platform-specific directories such as ~/.claude/projects, ~/.codex/sessions, ~/.hermes/sessions, and ~/.pi/agent.\n
  • Boundary markers: The audit script uses specific regex patterns to identify skill invocations (e.g., <command-name> tags) within the message content of the transcripts.\n
  • Capability inventory: The skill includes functionality to move directories and modify symbolic links on the local file system via scripts/archive.mjs.\n
  • Sanitization: The scripts use standard JSON parsing and regex-based extraction to process transcript data, ensuring that the content of the transcripts is treated as data for analysis rather than instructions.\n- [EXTERNAL_DOWNLOADS]: The documentation references installation via npx skills@latest, a standard method for managing agent skills, to fetch the skill from the author's repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 23, 2026, 05:09 AM
Security Audit — agent-trust-hub — aggressively-cleanup-skills