Checking ChEMBL for Structured SAR Data

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill performs network requests to www.ebi.ac.uk to access the ChEMBL database. This is a well-known scientific service provided by the European Bioinformatics Institute.- [COMMAND_EXECUTION]: The skill executes shell commands using curl and jq to fetch and parse JSON data. It also includes instructions to generate and run a Python helper script (check_chembl.py) for automated DOI validation.- [DATA_EXFILTRATION]: Extracted bioactivity data and document IDs are recorded in local project files, specifically SUMMARY.md and papers-reviewed.json. No sensitive information is sent to unauthorized external domains.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted data from an external API.
  • Ingestion points: JSON responses from the ChEMBL API ingested during workflow steps in SKILL.md.
  • Boundary markers: Absent; no instructions are provided to the agent to treat external content as untrusted.
  • Capability inventory: Shell execution via curl and python3, and file system write access for tracking files.
  • Sanitization: Structural parsing is performed using jq and json.json(), but no natural language sanitization is applied to the retrieved content.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 02:28 AM