academic-reference-finder

Pass

Audited by Gen Agent Trust Hub on Jul 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes a local Python script at .codex/skills/scripts/validate_references.py to validate metadata. The script is passed a JSON list of DOIs as a command-line argument.
  • [EXTERNAL_DOWNLOADS]: The skill queries established academic services including Crossref, arXiv, PubMed, Europe PMC, and Semantic Scholar to retrieve publication data. These are well-known and reputable bibliographic metadata providers.
  • [PROMPT_INJECTION]: The skill processes external research data, creating a surface for indirect prompt injection. Ingestion points: Metadata retrieved from external academic APIs. Boundary markers: None explicitly defined in the instructions to separate data from instructions. Capability inventory: Local script execution via python to validate references. Sanitization: No specific sanitization or filtering logic is described for the ingested metadata. The risk is considered low due to the specialized nature of the processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 26, 2026, 09:52 AM
Security Audit — agent-trust-hub — academic-reference-finder