brainstorming-partner
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The instructions in SKILL.md define a collaborative conversational framework for brainstorming. The skill focuses on role-play, structured dialogue phases, and opinionated feedback without involving dangerous system commands or file operations.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external data from the web using
WebSearchandWebFetch(SKILL.md) to ground its suggestions in real-world data. This presents a theoretical surface for indirect prompt injection if external websites contain malicious instructions targeting the LLM. - Ingestion points: Data enters the context via
WebSearchandWebFetchduring Phase 2 and Phase 3 of the brainstorming flow. - Boundary markers: The skill does not explicitly instruct the agent to use delimiters or ignore instructions found within search results.
- Capability inventory: The skill is limited to web search, fetching content, and conversation. It does not possess file system write access or arbitrary command execution capabilities.
- Sanitization: There are no instructions for sanitizing or filtering retrieved web content.
Audit Metadata