deep-research
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a research workflow that ingests data from external sources, which could contain malicious instructions designed to influence the agent.
- Ingestion points: External data enters the context through 'WebSearch' and 'WebFetch' tools mentioned in the Research Protocol section (SKILL.md).
- Boundary markers: The instructions lack explicit delimiting or escaping for external data, providing no clear separation between instructions and retrieved information.
- Capability inventory: The skill facilitates web-based data collection, synthesis, and reporting.
- Sanitization: No specific sanitization, filtering, or validation of the fetched content is defined in the protocol.
Audit Metadata