skills/jimliu/science-skills/chai1/Gen Agent Trust Hub

chai1

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads approximately 5 GB of model weights and auxiliary models (such as ESM2-3B) from Chai Discovery during the first execution. This is standard behavior for high-parameter foundation models.
  • [SAFE]: The skill clearly documents its data privacy implications, specifically noting that using the --use-msa-server flag sends biological sequences to the public ColabFold MSA server (api.colabfold.com).
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 01:21 AM
Security Audit — agent-trust-hub — chai1