sciverse-academic-retrieval

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a Model Context Protocol (MCP) client to interact with the SciVerse academic retrieval service. All operations are aligned with its stated purpose of scientific research.
  • [EXTERNAL_DOWNLOADS]: The skill connects to 'https://scp.intern-ai.org.cn' to perform search and retrieval operations. This domain is managed by the skill's authoring organization (OpenDataLab).
  • [DATA_EXFILTRATION]: The client requires an API key ('SCP-HUB-API-KEY') which is transmitted in the HTTP headers to the official gateway for authentication. This is standard practice for authenticated API access.
  • [COMMAND_EXECUTION]: The provided Python implementation includes logic to save retrieved figure/table data to the local file system using 'pathlib'. The 'get_resource' tool includes validation to prevent directory traversal by rejecting filenames with '..' or leading slashes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 04:18 PM
Security Audit — agent-trust-hub — sciverse-academic-retrieval