academic-research-hub

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches academic paper metadata and PDF documents from established repositories including arXiv and Semantic Scholar. These are trusted academic sources.
  • [DATA_EXFILTRATION]: Saves search results and downloaded papers to the local file system. This functionality is intended for organizing research data and does not involve sensitive user credentials.
  • [PROMPT_INJECTION]: The skill processes untrusted content such as paper abstracts and titles from external APIs, which represents an indirect prompt injection surface.
  • Ingestion points: API responses from arXiv, PubMed, and Semantic Scholar processed in scripts/research.py.
  • Boundary markers: None present in the output formatting logic.
  • Capability inventory: The script has file system write access for saving results and PDFs.
  • Sanitization: No specific sanitization or filtering is applied to ingested text before it is returned to the agent context.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 10:45 AM
Security Audit — agent-trust-hub — academic-research-hub