harness-creator-v3

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides tools for repository scaffolding and auditing. All analyzed scripts and templates perform legitimate development tasks such as file creation and structural validation.\n- [SAFE]: Instructions included in templates (AGENTS.md) and patterns promote agent safety, such as enforcing verification gates and defining security carve-outs that must not be simplified away (e.g., input validation and data loss prevention).\n- [SAFE]: Network operations are limited to standard package manager commands (npm install, pip install, etc.) in initialization scripts, which is expected behavior for the declared purpose of setting up a project environment.\n- [SAFE]: No obfuscation, hidden logic, or persistence mechanisms were found in the scripts, templates, or markdown documentation. The use of dynamic execution is limited to fixed Node.js snippets for probing project metadata (e.g., package.json scripts).
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 07:04 AM