error-handling
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration were detected. The skill is purely instructional and focused on error recovery and team management.- [COMMAND_EXECUTION]: The skill demonstrates how to configure platform hooks to execute local validation scripts and provides CLI commands (e.g., jq, cat, tmux) for inspecting the agent's internal configuration and task state. These are standard procedures for the intended debugging use case and are restricted to the agent's internal directory (~/.claude/).
Audit Metadata