{
  "name": "ares-general-v0",
  "purpose": "Small, licensed general-language pretraining experiment; not a production corpus.",
  "sources": [{"id":"tinystories","provider":"Hugging Face dataset hub","dataset":"roneneldan/TinyStories","split":"train","expected_field":"text","licence_review_required":true,"notes":"Streaming sample only. Confirm the current dataset-card licence and terms before every run."}],
  "prohibited": ["private conversations", "credentials", "unlicensed copyrighted collections", "personal data without documented lawful basis"],
  "approval_required": true
}