jk-literature-search

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads content from legitimate scientific metadata and publication repositories.
  • Fetches publication metadata from eutils.ncbi.nlm.nih.gov, api.crossref.org, and api.openalex.org.
  • Downloads Open Access PDFs from the official PMC S3 bucket (pmc-oa-opendata.s3.us-east-1.amazonaws.com) and major publisher domains (Elsevier, Nature, Wiley, etc.) via DOI redirection.
  • Implementation includes validation checks to ensure downloaded files match the expected PDF format.
  • [COMMAND_EXECUTION]: The skill's primary functionality is implemented through bundled Python scripts meant to be executed by the agent.
  • Logic for searching, ranking, and downloading is encapsulated in standalone scripts like literature_search.py and publication_download.py.
  • These scripts perform standard file system operations to organize and save search results.
  • [SAFE]: Comprehensive analysis confirms the skill follows security best practices for its intended domain.
  • Environment variables for NCBI API keys and user identity are handled using standard .env configuration patterns.
  • Dynamic imports observed in the scripts (requests) use hardcoded module names, presenting no risk of path injection or arbitrary code loading.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 12:22 PM
Security Audit — agent-trust-hub — jk-literature-search