research-proposal
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill operates as a specialized academic writing assistant with a controlled workflow including human-in-the-loop confirmation for research outlines before content generation.
- [SAFE]: External data ingestion is limited to legitimate academic sources (WebSearch for reviews/trends and Zotero for the user's own library) via declared platform tools. There is no evidence of the agent using these tools to exfiltrate sensitive information.
- [SAFE]: No obfuscation, prompt injection, or remote code execution patterns were detected. Reference files and templates are purely instructional and adhere to professional academic standards.
- [SAFE]: Tool usage for file management (Read, Write, Edit, Glob, Grep) is correctly scoped to the generation and management of the research proposal documents within the user's workspace.
Audit Metadata