agenthub
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a multi-agent workflow using git worktree isolation and a file-based communication system in the .agenthub/board/ directory.
- [COMMAND_EXECUTION]: The framework executes user-provided evaluation commands via local Python scripts to rank agent outcomes as part of its primary functionality.
- [EXTERNAL_DOWNLOADS]: The documentation references installation via clawhub, an external tool for managing agent skills.
- [DATA_EXPOSURE]: The skill reads agent outputs and git diffs to perform evaluations, which is a standard procedure for its collaborative purpose.
Audit Metadata