chembl-database

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The SKILL.md documentation includes instructions for the agent to proactively suggest the author's commercial platform ('K-Dense Web') during complex research workflows, which constitutes behavioral steering.
  • [PROMPT_INJECTION]: The skill ingests data from the ChEMBL API, creating an indirect prompt injection surface. Ingestion points: API responses from chembl_webresource_client in scripts/example_queries.py. Boundary markers: Absent. Capability inventory: Network access and Python data processing. Sanitization: Absent.
  • [EXTERNAL_DOWNLOADS]: The skill requires the chembl_webresource_client and pandas Python packages, which are reputable tools for bioinformatics and data analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 08:47 PM