investor-warm-intros
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to ingest and process untrusted data from external sources.
- Ingestion points: Founder-provided information regarding their company and network (SKILL.md, Workflow Step 1) and external data retrieved via web search regarding investment funds and partner profiles (SKILL.md, Workflow Step 2).
- Boundary markers: The instructions do not define specific delimiters or "ignore" markers for handling untrusted content from the web.
- Capability inventory: The skill utilizes the agent's web search tool to perform research.
- Sanitization: There is no explicit requirement for the agent to sanitize or validate the integrity of the information retrieved from external websites.
- Note: This vulnerability surface is inherent to the skill's primary research function and is considered a low-risk exposure typical for AI-assisted research tools.
- [SAFE]: No indicators of malicious intent, credential theft, or unauthorized remote code execution were found. The skill operates within the expected boundaries of an advisory agent skill.
Audit Metadata