tooluniverse-literature-deep-research

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill utilizes a large set of platform-authorized tools to access well-known academic services such as PubMed, ArXiv, UniProt, and Semantic Scholar. All external data retrieval is performed through these vetted interfaces.
  • [SAFE]: No malicious patterns, including prompt injection, data exfiltration, or obfuscation, were detected in the skill instructions or reference documentation.
  • [SAFE]: The skill lacks executable code (scripts), relying entirely on structured markdown instructions and platform tools, which significantly reduces the potential attack surface.
  • [SAFE]: The instructions for using the 'advanced_literature_search_agent' correctly identify external dependency requirements (Azure OpenAI key) as configuration details rather than exposing sensitive credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 02:30 AM