arbor

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The core functionality of the skill is managed by a transparent Python script (scripts/tree.py) that handles state management through local JSON files. No malicious patterns, unauthorized file access, or credential harvesting were detected.
  • [EXTERNAL_DOWNLOADS]: The skill references an external GitHub repository (github.com/RUC-NLPIR/Arbor) as a standalone alternative tool. This is a well-known research repository and the reference is documented neutrally for informational purposes.
  • [COMMAND_EXECUTION]: The skill coordinates the execution of evaluators and artifact modifications. These actions are performed using isolated subagents in separate git worktrees, which is a standard and secure practice for autonomous experimentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 03:28 AM
Security Audit — agent-trust-hub — arbor