agent-builder

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The documentation focuses on agent design theory and architectural patterns. No malicious instructions or safety bypasses were found.
  • [COMMAND_EXECUTION]: The skill references a local scaffolding script (scripts/init_agent.py) and implementation templates (references/minimal-agent.py), which are standard resources for a development-focused tool.
  • [DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or unauthorized data access patterns are present in the provided files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 12:24 PM
Security Audit — agent-trust-hub — agent-builder