agent-hub
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a orchestration layer for parallel agents using native Git worktrees for isolation. It uses standard file-based communication (markdown/YAML) and provides tools for session management, result ranking, and merging. No malicious patterns, prompt injections, or unauthorized data exfiltration were detected. Internal scripts (hub_init.py, board_manager.py, etc.) follow the stated purpose of managing the agent collaboration lifecycle.
Audit Metadata