Claude Code Guide

The complete guide to Claude Code setup (Opus 4.6, Sonnet 4.6, Haiku 4.5). 1M token context window. 100+ hours saved. 25 hook events. Agent teams and task management. Production-tested patterns for skills, hooks, and MCP integration.

View the Project on GitHub ytrofr/claude-code-guide

Claude Code Guide

The complete guide to Claude Code setup, skills, hooks, and MCP integration.

GitHub stars License: MIT Documentation

Production-tested Claude Code patterns. 240+ documented patterns, 100+ hours saved per year.

Models: Opus 4.6 Sonnet 4.6 Haiku 4.5 – 1M token context window – 25 hook events – Agent teams & task management

Install Best Practices (Any Project)

One command to install production-tested best practices into any project:

# Core install (rules + best practices doc)
curl -sL https://raw.githubusercontent.com/ytrofr/claude-code-guide/master/install.sh | bash

# Full install (rules + skills + commands + all 19 rules)
git clone https://github.com/ytrofr/claude-code-guide.git
cd claude-code-guide
./install.sh --full /path/to/your-project

# Everything including hooks + settings.json
./install.sh --with-hooks /path/to/your-project

# Install globally (applies to ALL your projects)
./install.sh --global

# Update to latest version
bash .claude/best-practices/update.sh

Installation Tiers

Tier Command What’s Installed
Core ./install.sh 6 rules + BEST-PRACTICES.md + CLAUDE.md import
Full --full Core + 19 rules + 3 skills + 5 slash commands
With Hooks --with-hooks Full + 4 hooks + settings.json

Component Flags (composable)

Flag What it Installs
--skills 3 starter skills to ~/.claude/skills/ (global, auto-discovered)
--commands 5 slash commands (/session-start, /advise, etc.)
--hooks 4 hook scripts + settings.json (formatting, skill detection)
--all-rules All 19 rules across 9 categories (global, planning, quality, etc.)

Claude Code will immediately apply all best practices in every session after installation.

See install.sh for all options including --update, --uninstall, --global.

Why This Guide?

Metric Result
Context Window 1M tokens
Models Opus 4.6, Sonnet 4.6, Haiku 4.5
Time Saved 100+ hours/year
Hooks 25 events automated
Skills 240+ native loading
Token Savings 47-70% per branch
Production Skills 240+ documented
MCP Integrations 13 servers, 70+ tools

Source: 14+ months of production use with 97 components validated.

Quick Start (30 minutes)

# 1. Clone the template
cp -r template/ ~/my-project/.claude
cd ~/my-project

# 2. Validate structure
./scripts/validate-setup.sh

# 3. Add skills (Claude Code discovers them automatically)
cp -r template/.claude/skills/ ~/.claude/skills/

# 4. Configure MCP servers
cp .claude/mcp_servers.json.template .claude/mcp_servers.json
# Add your GitHub token

# 5. Start Claude Code
claude
Detailed walkthrough: Quick Start Guide Skills System

Frequently Asked Questions

What is Claude Code?

Claude Code is Anthropic’s official CLI for AI-powered coding assistance, powered by Opus 4.6, Sonnet 4.6, and Haiku 4.5 with a 1M token context window. It provides an interactive terminal experience where Claude can read files, write code, run commands, manage tasks, and coordinate agent teams. Claude Code understands your project context through CLAUDE.md files and can be extended with hooks, skills, and MCP servers.

How do I set up Claude Code?

Install via the official installer: curl -fsSL https://claude.ai/install.sh | sh (or claude update if already installed). Create a CLAUDE.md file in your project root with project-specific instructions. Optionally add hooks in .claude/hooks/ for automation, skills in ~/.claude/skills/ for reusable workflows, and MCP servers for database/API access. Our template provides all these pre-configured.

What are Claude Code hooks?

Hooks are shell scripts that run automatically at specific points in the Claude Code lifecycle. There are 25 hook events (PreToolUse, PostToolUse, SessionStart, SessionEnd, etc.) that can validate inputs, block dangerous operations, auto-format code, and run background analytics. See our Complete Hooks Guide.

What is MCP integration?

MCP (Model Context Protocol) extends Claude Code with external tools. Connect to PostgreSQL databases, GitHub repositories, memory systems, and APIs. This guide covers 13 MCP servers with 70+ tools including PostgreSQL, GitHub, Perplexity, Playwright, and Basic Memory for persistent knowledge storage.

How do Claude Code skills work?

Skills are Markdown files with YAML frontmatter (name: and description: with “Use when…” clause). Claude Code natively discovers all skills from ~/.claude/skills/ and matches them to queries using the description field. No custom hooks needed – skills are built into Claude Code. Our system runs 240+ production skills with native activation.

What is the memory bank?

The memory bank is a hierarchical knowledge system using a 4-tier structure: always (auto-loaded), learned (patterns), ondemand (blueprints), and reference (archives). It stores project context, decisions, and patterns for efficient token usage. Properly configured, it provides 34-62% token reduction while maintaining full context access.

How much time does this save?

Based on production metrics: 100+ hours per year in developer time. Key achievements include 240+ proven patterns, 47-70% token savings per branch, 25 hook events for automation, and 80%+ research cost savings with Perplexity caching. ROI scales with project complexity.

Core Documentation

Getting Started

Claude Code Hooks

Skills System

MCP Integration

Context Optimization

Best Practices

Agents & Teams

Repository Structure

claude-code-guide/
├── install.sh               # One-command best practices installer
├── best-practices/          # Installable best practices package
│   ├── BEST-PRACTICES.md   # Universal best practices document
│   ├── rules/              # 6 universal rules (project-agnostic)
│   └── VERSION             # Version tracking for updates
├── docs/                    # Complete documentation
│   ├── quick-start.md      # 30-minute setup
│   ├── skill-activation-system.md
│   └── guide/              # 59 detailed guides
├── template/                # Clone-and-go starter (full setup)
│   ├── .claude/            # Pre-configured setup
│   │   ├── CLAUDE.md       # Project context
│   │   ├── hooks/          # 9 automation scripts
│   │   ├── rules/          # 19 rules across 9 categories
│   │   └── skills/         # Starter skills
│   └── memory-bank/        # Knowledge hierarchy
├── skills-library/          # 20+ proven workflows
├── mcp-configs/             # MCP server configs
├── scripts/                 # Setup & validation
├── examples/                # Real-world examples
└── web/                     # Interactive checklist

Phased Implementation

Phase 0: Minimal (30 min)

Phase 1: Essential (Week 1)

Phase 2: Productive (Week 2-3)

Phase 3: Advanced (Month 2+)

Key Features

What Makes This Different

Aspect This Guide
Production-Tested 240+ patterns from real systems
Evidence-Based Every claim backed by metrics
Team-Ready Shared vs personal configuration
Phased Approach Value in 30 minutes
Validation-First Scripts catch 90% of mistakes

Credits

Source Project: Production system (97 components) Research: Anthropic Claude 4 Best Practices + Sionic AI Skills Framework Marketplace: wshobson/agents - 273 components Official Docs: code.claude.com/docs Created: December 2024 Updated: April 2026

License

MIT License - See LICENSE

Getting Started

Core Systems

Advanced Topics

Built with lessons from 14+ months of production Claude Code usage.