A comprehensive guide for prompt engineering containing papers, guides, lessons, notebooks, and resources on prompt engineering, context engineering, RAG, and AI Agents.
One-Minute Overview#
This is a comprehensive resource collection on prompt engineering maintained by the DAIR.AI team. Whether you're a researcher or developer, this guide will help you understand the capabilities and limitations of large language models and learn how to design more effective prompts to optimize your AI interactions.
Core Value: Provides a systematic knowledge base for prompt engineering, from basic concepts to advanced techniques, helping users fully unlock the potential of AI models.
Quick Start#
Installation Difficulty: Medium - Requires Node.js environment and pnpm package manager
# Install dependencies
pnpm i next react react-dom nextra nextra-theme-docs
# Run development server
pnpm dev
Is this suitable for me?
- ✅ Researchers: Need to understand LLM capabilities and limitations with latest papers
- ✅ AI Developers: Looking for practical prompt engineering techniques and best practices
- ✅ AI Learners: Want to systematically learn about prompt engineering, RAG, and AI agents
- ❌ Complete Beginners: Should first understand basic AI concepts before using this resource
Core Capabilities (Optional)#
1. Systematic Prompt Engineering Tutorials - From Basic to Advanced#
- Covers various prompting techniques including zero-shot, few-shot, chain-of-thought Actual Value: Helps users select the most suitable prompting methods for different scenarios
2. Rich Model-Specific Guides - Optimized for Different Models#
- Provides specialized guides for mainstream models like ChatGPT, GPT-4, Gemini, LLaMA Actual Value: Helps users understand specific model characteristics and best practices, avoiding suboptimal results with generic prompts
3. Practical Applications and Scenarios - Theory Meets Practice#
- Includes real-world application cases like code generation, dataset creation, Q&A systems Actual Value: Transforms theoretical knowledge into practical productivity, helping users solve real problems
4. Risk and Bias Research - Safe AI Usage#
- Explores risk issues like adversarial prompting, factuality, and biases Actual Value: Helps users identify and avoid potential pitfalls in AI usage, ensuring responsible technology application
Tech Stack and Integration (Optional)#
Development Languages: MDX, Jupyter Notebook, TypeScript Main Dependencies: Next.js, React, Nextra Integration Method: Web Guide / Locally runnable knowledge base
Maintenance Status (Optional)#
- Development Activity: High - Project has 1584 commits with continuous updates
- Recent Updates: Active - Regularly adds new prompting techniques and model guides
- Community Response: Active - Has Discord community and multiple social media channels
Commercial and License (Optional)#
License: MIT
- ✅ Commercial Use: Allowed
- ✅ Modification: Allowed
- ⚠️ Restrictions: Must include author attribution information
Documentation and Learning Resources (Optional)#
- Documentation Quality: Comprehensive - Includes full guides, examples, and tutorials
- Official Documentation: https://www.promptingguide.ai/
- Example Code: Provides complete Jupyter notebooks and code examples
- Learning Resources: Offers video lectures, slides, and accompanying courses