brainstorm-ideas-existing
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data which could contain hidden instructions intended to influence the agent's behavior.
- Ingestion points: The instructions in
SKILL.mddirect the agent to read user-provided files (research data, personas) and perform web searches on user-provided product URLs. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore or isolate potential commands embedded within the external research data or web content.
- Capability inventory: The skill has the capability to read local files, perform web searches, and write output files to the user's workspace.
- Sanitization: The skill lacks logic to sanitize, validate, or filter content retrieved from external URLs or files before processing it into the final output.
Audit Metadata