Overview
A comprehensive repository of prompt engineering patterns tailored for Claude’s high-context window and constitutional AI training. Covers both system and user-level prompt optimization.
Features
- JSON Mode Patterns: reliable ways to force Claude to output structured data.
- Role Mastery: Deeply defined personas for code review, architecture, and scientific analysis.
- Technique Library: Examples of Chain-of-Thought, Few-Shot, and Prompt Chaining.
Use Cases
- Designing high-quality system prompts for custom Claude agents.
- Improving the reliability of automated data extraction tasks.
- Optimizing context injection for long-running coding sessions.