---

🎯 Claude Prompt Engineering Guide

🚀 The definitive guide to writing professional Claude Standard prompts for Opus, Sonnet, and Haiku models with comprehensive coverage of MCP, Skills, Superpowers, and advanced prompt engineering techniques.

---

📖 Table of Contents

• Overview
• Features
• Quick Start
• Skills Collection
• Core Content
• Documentation Structure
• Key Sections
• Examples & Templates
• Contributing
• License
• Acknowledgments

---

🌟 Overview

This comprehensive guide synthesizes Anthropic's official best practices with real-world prompt engineering techniques for Claude 4.x models. Whether you're using Claude through the web interface, desktop app, Claude Code CLI, or the API, this guide provides proven patterns and frameworks for extracting maximum value from Claude's capabilities.

Who Is This For?

• Developers building applications with Claude's API
• Prompt Engineers designing production prompts for teams
• AI Engineers integrating Claude into workflows
• Claude Code Users leveraging agentic capabilities
• Researchers exploring Claude's reasoning abilities
• Anyone wanting to master professional prompt engineering

Why This Matters

Claude 4.x models are extraordinarily capable, but extracting that capability requires structured prompting. This guide provides:

✅ Anthropic's 10-Component Framework — The official structure for professional prompts
✅ Claude 4.x Best Practices — Specific techniques for Opus, Sonnet, and Haiku models
✅ Advanced Techniques — XML tagging, chain of thought, extended thinking, and more
✅ Real-World Patterns — Code review, business analysis, research, document creation
✅ Tool Integration — MCP, Skills, Superpowers, and Perplexity integration
✅ Environment Guides — Optimal approaches for Claude.ai, Desktop, Code, and API

---

✨ Features

This guide includes:

• 📚 1000+ lines of comprehensive reference material
• 🏗️ Official 10-component prompt framework with detailed explanation
• 💡 5 advanced prompt patterns with complete examples
• 🛠️ Tool integration guides (MCP, Skills, Superpowers)
• 🎯 Environment-specific optimizations (web, desktop, CLI, API)
• 📋 Prompt templates (minimal and comprehensive)
• 🔍 Real-world use cases across multiple domains
• ⚙️ Model comparison chart (Opus vs Sonnet vs Haiku)
• 📊 Pricing and performance guide
• 🚀 Best practices for long-horizon reasoning
• 🧠 Chain of thought and extended thinking techniques
• 🔐 Security and prompt injection prevention

---

🚀 Quick Start

1. Read the Main Guide

Start with the comprehensive Claude Prompt Engineering Guide which covers:

• Claude's architecture and philosophy
• The 10-component framework
• Best practices for Claude 4.x
• Advanced techniques
• Complete pattern examples

2. Choose Your Environment

• Using Claude.ai? → Read Claude.ai Optimization Guide
• Using Claude Desktop? → Read MCP Integration Guide
• Using Claude Code CLI? → Read Claude Code Guide
• Building with API? → Read API Integration Guide

3. Find Examples for Your Use Case

• Coding Tasks
• Research & Analysis
• Business Analysis
• Document Creation

4. Use a Template

Customize one of our prompt templates:

• Minimal Prompt Template — Quick projects
• Comprehensive Prompt Template — Complex tasks

5. Explore Claude Skills

Discover reusable skill packages in our growing collection:

• Skills Directory — Browse available skills and contribute your own
• Skill Template — Example feedback analyzer skill
• Learn to create skills — Full documentation in skills/README.md

---

📦 Skills Collection

What Are Claude Skills?

Claude Skills are modular, reusable task packages that extend Claude's capabilities with domain-specific knowledge, procedures, and workflows. They're designed to be:

• ✅ Modular — Self-contained, focused on specific tasks
• ✅ Reusable — Used across different conversations and projects
• ✅ Composable — Multiple skills can work together seamlessly
• ✅ Discoverable — Claude automatically identifies relevant skills
• ✅ Efficient — Progressive disclosure prevents context bloat

Available Skills

Our comprehensive collection includes 22 production-ready skills:

Web Development: NextJS App Router, Tailwind Design System, NextAuth, API Development
Infrastructure: AWS, GCP, Neon Serverless, Prisma ORM
Testing: Vitest, Playwright E2E, Code Review, Testing Frameworks
DevOps: Vercel Deployment, Database Migrations, Monitoring & Logging, Git Workflow
Standards: TypeScript, Performance, SEO, Security, Accessibility, Feedback Analysis

→ View All 22 Skills

Quick Links

• 📚 Full Skills Documentation — Learn everything about skills
• 🛠️ How to Create Skills — Step-by-step guide
• 📋 Skill Template — Use as a starting point
• 🤝 Contribute Your Skill — Share with the community

Why Use Skills?

Skills help you:

• 📚 Standardize processes across your team
• 🎯 Ensure consistency in outputs and workflows
• ⏱️ Save time with pre-built procedures
• 🔧 Customize Claude for your specific domain
• 📈 Improve quality through proven patterns

Getting Started

1. Browse available skills in skills/examples/
2. Copy a skill to use in your conversations
3. Reference the skill in your prompts: "Use the [Skill Name] to..."
4. Create your own skills following our template
5. Contribute your skills back to the community

---

📚 Core Content

Claude Prompt Engineering Guide

The comprehensive reference document containing:

Section 1: Understanding Claude's Architecture

• Claude's character and philosophy
• Knowledge cutoff dates
• How Claude processes prompts

Section 2: Claude Models Overview

• Claude Opus 4.1 — Most powerful model
• Claude Sonnet 4.5 — Balanced performance and cost
• Claude Haiku 4.5 — Fast and efficient
• Pricing and performance comparison

Section 3: System Prompts vs User Prompts

• When to use system prompts
• When to use user prompts
• Best practices for each

Section 4: Anthropic's Official Prompt Structure

• The 10-Component Framework (official structure)
• Component explanations and examples
• Why this structure works

Section 5: Claude 4.x Best Practices

• Be explicit with instructions
• Add context to improve performance
• Long-horizon reasoning techniques
• State tracking best practices
• Tool usage patterns
• Output formatting control
• Parallel tool calling
• Research approaches
• Avoiding hallucinations

Section 6: Advanced Techniques

• XML tags for structure
• Chain of thought prompting
• Extended thinking
• Prompt chaining
• Role prompting

Section 7: Tools, MCP, Skills & Superpowers

• Model Context Protocol (MCP)
• MCP Filesystem Server
• Claude Skills system
• Superpowers plugin by obra
• Perplexity MCP integration

Section 8: Prompt Engineering for Different Environments

• Claude.ai web interface
• Claude Desktop app
• Claude Code (CLI/VS Code)
• Claude API (direct integration)

Section 9: Common Patterns & Examples

• Pattern 1: Technical Code Review
• Pattern 2: Business Analysis with Data
• Pattern 3: Long-Horizon Coding Tasks
• Pattern 4: Research and Synthesis
• Pattern 5: Document Creation with Skills

Section 10: Quick Reference Card

• Minimal prompt template
• Comprehensive prompt template
• Quick checks checklist

---

📖 Documentation Structure

claude-prompt-engineering-guide/
├── README.md                          # This file
├── Claude-Prompt-Guide.md             # Main comprehensive guide
├── LICENSE                            # MIT License
├── CONTRIBUTING.md                    # Contribution guidelines
├── CHANGELOG.md                       # Version history
├── .gitignore                         # Git ignore rules
│
├── docs/                              # Additional documentation
│   ├── quick-start.md                # Getting started guide
│   ├── mcp-integration.md            # MCP setup and usage
│   ├── skills-guide.md               # Skills documentation
│   ├── superpowers-guide.md          # Superpowers plugin guide
│   ├── api-guide.md                  # API integration guide
│   ├── claude-code-guide.md