wealth-structure
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied prompts regarding career and financial decisions, which serves as a surface for indirect instructions.
- Ingestion points: User queries containing keywords like 'equity', 'leverage', or '致富' (get rich) are processed through the methodology outlined in SKILL.md.
- Boundary markers: The skill includes a 'Boundary' section that instructs the agent not to provide specific investment advice or use the skill for survival-level financial decisions.
- Capability inventory: The skill configuration includes no code, no tool permissions, and no network or file system capabilities, which effectively mitigates the impact of potential injections.
- Sanitization: There are no explicit routines for filtering or sanitizing incoming user text before it is analyzed by the agent.
Audit Metadata