fengge-capital
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process arbitrary user-provided topics related to finance, stocks, and elite culture to generate persona-driven responses.
- Ingestion points: User input in the form of financial topics or specific prompts (e.g., "帮我用资本峰锐评") as defined in
SKILL.md. - Boundary markers: The skill does not implement specific boundary delimiters or safety instructions to isolate user-provided data from the agent's internal logic.
- Capability inventory: The skill consists entirely of instructional text and does not utilize any tools, file system operations, network requests, or code execution capabilities.
- Sanitization: No explicit input validation or output sanitization mechanisms are defined.
- [METADATA_POISONING]: The skill contains hypothetical future-dated references (e.g., May 2026) and mentions real-world financial figures (e.g., Duan Yongping, Nvidia) to establish the "Capital Feng" persona. These elements are consistent with its intended role-play purpose and do not represent a security risk.
Audit Metadata