content-research-writer
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to ingest and process untrusted data from external sources during its research phase.
- Ingestion points: The skill retrieves external web content during research tasks and processes user-provided drafts in the local environment.
- Capability inventory: The skill performs filesystem writes (creating/updating drafts) and network operations (web search for research).
- Boundary markers: The instructions direct the agent to use Markdown formatting (blockquotes and headers) to delineate its feedback and research findings from the user's content.
- Sanitization: No explicit sanitization or filtering of external research content is defined in the instructions, relying on the model's inherent ability to parse relevant information.
Audit Metadata