cancel

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use terminal commands to manage local state files and terminate processes.
  • Ingestion points: Reads state from .omc/state/*.json and team configuration from ~/.claude/teams/*/config.json.
  • Boundary markers: None explicitly mentioned in the instructions for data reading.
  • Capability inventory: Uses rm, find, cat, jq, ps, tmux, and kill to perform cleanup tasks.
  • Sanitization: Relies on internal state tools and agent-managed paths to ensure cleanup is restricted to specific OMC directories.
  • [COMMAND_EXECUTION]: Execution of a local maintenance script to clean up background processes.
  • Evidence: The skill calls node "${CLAUDE_PLUGIN_ROOT}/scripts/cleanup-orphans.mjs" to identify and terminate lingering agent processes, using local environment variables for the path.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 12:19 PM