scholar-ideation
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs legitimate academic research tasks using agent tools (WebSearch, WebFetch, Write) as intended for its primary purpose.
- [SAFE]: No evidence of prompt injection, obfuscation, or unauthorized data exfiltration was detected in the skill instructions or associated phase documents.
- [SAFE]: External communication is restricted to well-known academic databases (arXiv, Google Scholar) and user-configured MCP tools for Zotero integration.
- [SAFE]: The skill incorporates human-in-the-loop checkpoints via AskUserQuestion, ensuring the user reviews and confirms research directions and keywords.
Audit Metadata