Getting Started
Go from zero to productive with Claude Code in 10 minutes.
Step 0: Create a Claude Account
If you don’t have one yet, sign up at claude.ai and choose a plan. See the Account Setup guide for plan comparisons and recommendations.
Step 1: Install the Playbook (2 minutes)
Option A — Interactive installer:
curl -sL https://raw.githubusercontent.com/ao92265/claude-code-playbook/main/scripts/install.sh | bash
Option B — Manual setup:
git clone https://github.com/ao92265/claude-code-playbook.git
cd claude-code-playbook
# Copy a template to your project
cp templates/CLAUDE.md /path/to/your/project/CLAUDE.md
# Install skills globally
cp -r skills/check-env ~/.claude/skills/
cp -r skills/deploy ~/.claude/skills/
cp -r skills/handoff ~/.claude/skills/
# Install hooks
mkdir -p ~/.claude/hooks
cp hooks/ts-check.sh ~/.claude/hooks/
chmod +x ~/.claude/hooks/ts-check.sh
Step 2: Customize Your CLAUDE.md (3 minutes)
Open the CLAUDE.md you just copied and edit each section:
- Project Basics — Set your language, framework, and module system
- Testing — Specify your test runner and testing conventions
- Verification — Define what “done” means (which commands to run)
- Git & GitHub — Keep the safety rules as-is
Delete sections that don’t apply. Every line should prevent a real problem.
Step 3: Configure Hooks (2 minutes)
Add to ~/.claude/settings.json:
{
"hooks": {
"PostToolUse": [
{
"matcher": "Edit|Write",
"hooks": [
{
"type": "command",
"command": "~/.claude/hooks/ts-check.sh",
"timeout": 30,
"statusMessage": "Checking TypeScript types..."
}
]
}
]
}
}
For the complete 7-hook setup, see config/hooks-example.json.
Step 4: Your First Session (3 minutes)
Start Claude Code and try:
> /check-env
This validates your environment — ports, Docker, credentials, Node.js. Fix any issues it finds.
Then try a real task:
> Fix the failing test in src/auth/login.test.ts.
The error is: "Expected 200, received 401".
Only modify files in src/auth/. Don't change tests.
Notice the pattern: specific error + scope constraint + clear goal.
What to Learn Next
| Goal | Resource |
|---|---|
| Master the core workflow | Power User Guide |
| Improve your prompts | Prompt Patterns |
| Avoid common mistakes | Anti-Patterns |
| Choose the right skill | Workflow Decision Tree |
| Pick the right model | Model Comparison |
| Set up MCP servers | MCP Servers Guide |
| See real examples | Example Sessions |
The 5 Rules That Matter Most
If you remember nothing else from this playbook:
- One task per session. Multi-task sessions degrade quality.
- Paste real errors. Don’t describe — paste the stack trace.
- Scope-lock your prompts. “Only modify
src/auth/” prevents drift. - Verify with real tests. “Looks correct” ≠ “passes tests”.
- Start fresh above 80% context. Use
/handoffto preserve context for the next session.