repository-harvesting-guide

Pass

Audited by Gen Agent Trust Hub on Apr 2, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is strictly educational and provides templates for interacting with well-known academic metadata providers using the OAI-PMH protocol.
  • [COMMAND_EXECUTION]: Includes Python code snippets that use the 'requests' and 'xml.etree.ElementTree' libraries for standard HTTP operations and XML parsing. There is no evidence of arbitrary command execution or unsafe subprocess calls.
  • [EXTERNAL_DOWNLOADS]: The skill mentions and provides code to interact with reputable academic repositories such as arXiv, PubMed Central, and Europeana. These are well-known, trusted services in the research community.
  • [DATA_EXFILTRATION]: No patterns of sensitive data access (credentials, SSH keys, or environment variables) were detected. The network operations are limited to fetching public metadata from specified repository endpoints.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 2, 2026, 03:09 PM
Security Audit — agent-trust-hub — repository-harvesting-guide