deep-research
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses built-in orchestration patterns to manage research tasks. It utilizes native
WebSearchandWebFetchtools for their intended purposes. - [SAFE]: The workflow includes explicit requirements for evidence triangulation, cross-checking claims across multiple sources, and mandatory citations, which promotes accurate and verifiable output.
- [SAFE]: No sensitive file access, network exfiltration, or hardcoded credentials were detected in the skill's instructions or metadata.
- [SAFE]: The instructions for parallel sub-agent execution (
ResearchScout) are logical and scope-limited, focusing only on information retrieval and summarization.
Audit Metadata