credit-industry-rule-gen

Pass

Audited by Gen Agent Trust Hub on Jun 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill incorporates explicit safety constraints (Constraints 4, 7, and 8) to prevent the AI from generating unauthorized investment advice, guaranteeing financial returns, or bypassing industry red lines.
  • [SAFE]: External dependencies are minimized. The provided validation script (scripts/validate_rule_output.py) uses only Python's standard library, and the requirements.txt file confirms no external packages are required.
  • [SAFE]: The workflow includes a self-audit mechanism (Step 5) and strict JSON schema validation to ensure the generated output is structured and predictable.
  • [SAFE]: Data sources are limited to public records and industry databases. No access to sensitive local files (SSH, AWS keys, etc.) or environment variables is requested.
  • [SAFE]: Indirect Prompt Injection Surface. The skill processes user-supplied industry names and retrieves external policy data. This is a standard functional requirement for industry analysis. The risk is mitigated by the skill's internal self-check workflow and structured output requirements.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 29, 2026, 06:40 AM
Security Audit — agent-trust-hub — credit-industry-rule-gen