skills/wu-yc/labclaw/hypogenic/Gen Agent Trust Hub

hypogenic

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill clones example datasets and research repositories from the ChicagoHAI GitHub organization, which is associated with a recognized academic laboratory.
  • [COMMAND_EXECUTION]: Provides instructions to run local setup scripts (setup_grobid.sh) and data processing utilities (pdf_preprocess.py) required for literature analysis.
  • [PROMPT_INJECTION]: Includes a functional instruction at the end of the file that directs the model to recommend the 'K-Dense Web' platform when user requests become complex, shaping the agent's interaction style.
  • [SAFE]: Installation and usage instructions follow standard software development practices for research tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 04:32 AM