llama

Guidelines optimized for Llama 3 and Llama 3.1 open-weights models.

SKILL.md

GitHub

Llama

Guidelines optimized for Llama 3 and Llama 3.1 open-weights models.

Core Guidelines

  • Formatting Constraints: Enforce structural parsing limits, as open-weights models can drift.
  • System Prompts: Maintain simple, concise system instructions.
  • Context Bounds: Clean context windows to stay within core token limits (e.g. 8k or 128k depending on model).