quant-credit-risk
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, data exfiltration attempts, or prompt injection vectors were identified. The content is strictly related to credit risk modeling terminology. The use of non-Latin characters "迁移矩阵" refers to "Migration Matrix," a standard financial tool for tracking credit rating changes, and is not an obfuscation attempt.
- [NO_CODE]: The skill does not contain any scripts, executables, or code snippets. It operates purely as an instructional framework for the AI agent to produce text-based analysis.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied problem descriptions, but the overall risk is assessed as safe because the skill has no functional capabilities to interact with the system or network.
- Ingestion points: User-provided problem descriptions and credit inputs in SKILL.md.
- Boundary markers: None present.
- Capability inventory: No executable capabilities, network operations, or file system access identified; restricted to markdown artifact generation.
- Sanitization: None present.
Audit Metadata