agent-creator
Pass
Audited by Gen Agent Trust Hub on May 20, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches occupational alignment data from authoritative domains such as bls.gov and mymajors.com to ground agent definitions in industry standards.
- [COMMAND_EXECUTION]: Uses local scripts (scripts/main.cjs) and node sub-processes to automate registry regeneration and routing table updates within the workspace.
- [PROMPT_INJECTION]: Employs an internal 'Security Review Gate' that proactively scans external fetches for prompt injection patterns, oversized content, and unauthorized tool invocations.
- [DATA_EXFILTRATION]: Updates internal memory files (.claude/context/memory/) and routing configuration (CLAUDE.md) to maintain persistent state across agent spawning sessions.
Audit Metadata