investor-letter-writer
Pass
Audited by Gen Agent Trust Hub on Nov 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user inputs and web search results to generate structured investment memos, creating a surface for potential indirect prompt injection attacks.
- Ingestion points: Processes user-provided company names, sectors, and data, as well as external content retrieved via web searches as described in SKILL.md and references/analysis_framework.md.
- Boundary markers: The instructions do not define clear delimiters or 'ignore embedded instructions' warnings for separating external data from instructions.
- Capability inventory: The agent can write files in DOCX, PDF, and PPTX formats and perform web searches.
- Sanitization: No evidence of sanitization, filtering, or validation of the ingested external content is mentioned.
Audit Metadata