A practical guide to Claude Code — from your first prompt to multi-agent automation, hooks, MCP, and team workflows. Built around clear mental models and real examples, not marketing.

npm install -g @anthropic-ai/claude-code

Who this is for: Developers using (or about to use) Claude Code. Beginners get a guided path; power users get depth on Skills, Hooks, MCP, and Agent Teams.

The four extension points in Claude Code, side by side:

💡 These four compose. Most polished workflows combine 2–3.

Fundamentals — What is Claude Code? · Setup · Prompt Engineering

Workflow extensions — Slash Commands · Skills · Hooks

Multi-agent & integration — Subagents · Agent Teams · MCP

Productivity & frameworks — Effort levels · Fast Mode · Super Claude · BMAD Method

Reference — Slash Command Cheatsheet · Effort Levels · FAQ · Updates & Deprecations · Further Reading

Claude Code is Anthropic's official CLI for working with Claude from your terminal. You point it at a project; it reads the code, plans, edits files, runs commands, and commits — all from the prompt line.

Three things it does that a chat UI can't:

• Reads your actual repo — not pasted snippets. Claude sees your file tree, runs grep, follows imports, and grounds answers in real context.
• Edits in place and runs your tests — diff-aware edits, then pytest/vitest/go test on the spot to verify the change.
• Composes with the rest of your stack — slash commands, hooks, sub-agents, MCP servers, and your normal git/shell workflow.

If you've used Copilot or Cursor, think of Claude Code as their "agent in your terminal" peer — same idea, different surface, no editor lock-in.

claude          # start a session in the current repo
> explain what this codebase does
> fix the failing test in src/api.test.ts
> open a PR with the changes

Claude Opus 4.7 is the current flagship in the Claude 4 family (May 2026) — sharper adaptive thinking, better long-context grounding, more reliable tool calling than Opus 4.6, at the same price. 200K context (1M beta via API), 128K max output.

Choosing a model — quick guide:

→ Full specs, capabilities, and pricing in docs/reference/models.md

⏱️ 5-minute setup. Get from zero to your first AI-assisted commit.

npm install -g @anthropic-ai/claude-code

Requires Node.js 18+. For other install methods (Homebrew, curl, native binary), see the official install guide.

claude

On first run, Claude Code opens a browser to sign in with your Anthropic account (Pro, Max, or API key all work). After that, you can re-authenticate any time with /auth login from inside Claude.

From any project directory:

cd ~/your-project
claude

Once Claude Code is running, try one of these:

• explain what this codebase does — Claude reads your repo and summarizes.
• add a README section about installation — generates content based on your project.
• find and fix the failing test in src/api.test.ts — diagnoses and edits in place.

/init

Creates a project-level instruction file that Claude reads on every session — your project's "house rules." More on this in Prompt Engineering Deep Dive.

This repo's own .claude/ directory is a working example of a fully-configured Claude Code project. Browse it as a reference for what a polished setup looks like:

💡 Next: Once you're comfortable with the basics, jump to Claude Skills to build reusable slash commands in 3 minutes.

📖 Claude Initialization Run the /init command to automatically generate a CLAUDE.md file. Your CLAUDE.md files become part of Claude's prompts, so they should be refined like any frequently used prompt. A common mistake is adding extensive content without iterating on its effectiveness. Take time to experiment and determine what produces the best instruction following from the model.

Versatile workflow for complex problems.

• Explore: Read relevant files/images/URLs; use subagents for verification. Do not code yet.
• Plan: Ask Claude to make a plan. Use "think", "think hard", "think harder", or "ultrathink" to nudge depth in the prompt — see Effort levels for the full reasoning dial. Optionally save the plan for future reference.
• Code: Implement the solution; verify reasonableness as you go.
• Commit: Commit results, create pull requests, update READMEs/changelogs.
• Claude has two default modes: Plan Mode and Accept Edits Mode. You can toggle between them using the Shift + Tab keys.
  • 
  • 

💡 Pro Tip: Research & planning first significantly improves performance for complex tasks.

Ideal for changes verifiable with unit/integration tests.

• Write Tests: Create tests based on expected inputs/outputs; mark as TDD.
• Run & Fail Tests: Confirm they fail; no implementation yet.
• Commit Tests: Commit once satisfied.
• Write Code: Implement code to pass tests; iterate with verification via subagents.
• Commit Code: Final commit after all tests pass.

🔹 Clear targets (tests, mocks) improve iteration efficiency.

• Provide screenshots or visual mocks.
• Implement code, take screenshots, iterate until outputs match mock.
• Commit once satisfied.

🔹 Iteration significantly improves output quality (2-3 rounds usually enough).

→ Full guide in docs/reference/effort-levels.md

Mental model: Effort is a behavioural dial, not a token budget — it shifts thinking depth, tool-call appetite, response length, and how persistently Claude pushes through multi-step work. Higher ≠ smarter; context quality often matters more.

5 levels exist (most users assume 4):

Current defaults (May 2026): Opus 4.7 → xhigh; Opus 4.6 + Sonnet 4.6 → high on every plan. (The "Pro/Max users on a nerfed medium default" lore was true ~March → late April 2026 but is fixed in Claude Code v2.1.117. Check yours with /effort.)

Setting it, in order of persistence:

# This turn only — adds an in-context cue (does not change API effort)
> ultrathink — design the migration strategy

# This session — slider with no args, level name with arg
/effort xhigh
/effort auto                    # reset to model default

# All sessions, low/medium/high/xhigh — settings.json
echo '{ "effortLevel": "high" }' > .claude/settings.json

# All sessions, max — only the env var works
export CLAUDE_CODE_EFFORT_LEVEL=max

⚠️ Three gotchas worth knowing:

• Anthropic's own guidance for Opus 4.7 max: "shows diminishing returns and is more prone to overthinking" on routine work. Don't default to it.
• "effortLevel": "max" in settings.json silently downgrades — only CLAUDE_CODE_EFFORT_LEVEL=max env var persists max.
• Context quality often beats more effort. If you're reaching for max on a task that shouldn't need it, ~80% of the time the fix is upstream — sharper CLAUDE.md, atomic plan, named files. Full breakdown →

💡 Pattern: plan-with-Opus / execute-with-Sonnet. Plan in Opus xHigh or Max; hand the atomic, zero-ambiguity plan to Sonnet at lower effort to execute. Sonnet follows clear plans without drift, so the cheap execution is reliable when the plan is sharp.

Claude Code ships ~30 built-in slash commands plus the ability to define your own as skills (markdown files in .claude/commands/). The two work together — built-ins for common operations, custom skills for your team's workflows.

→ Full slash-command reference in docs/reference/commands.md (~30 commands including /auth, /fast, /hooks, /mcp, /teleport, …)

Define a frequently-used prompt once as a markdown file, invoke it forever with /skill-name:

mkdir -p .claude/commands
echo "Analyze this code for performance issues and suggest optimizations:" \
  > .claude/commands/optimize.md

💡 Next level: custom slash commands and Skills are the same thing. Head to Claude Skills for the deep dive — built-in skills, the 7 custom skills in this repo, workflow recipes, and how to write your own.

~3 min read · Full guide in docs/skills.md →

Mental model: Skills package a workflow into a markdown file. Two flavors:

• Slash skills — .claude/commands/<name>.md, you invoke them with /<name>
• Agent Skills — .claude/skills/<name>/SKILL.md with YAML frontmatter; Claude auto-invokes when the description matches the task

⚠️ Security: Skills are executable instructions running with your shell permissions. Read every third-party skill before adding it — exactly like reviewing a shell script before sourcing it.

mkdir -p .claude/commands

cat > .claude/commands/analyze.md << 'EOF'
# Code Analysis

Analyze the current code for:
- Potential bugs and edge cases
- Performance optimizations
- Code quality improvements
- Security vulnerabilities

Provide specific, actionable recommendations.
EOF

claude       # then type: /analyze

That's it — a working slash skill. Promote it to an Agent Skill later by moving it to .claude/skills/analyze/SKILL.md and adding name/description frontmatter.

The full Skills guide in docs/skills.md covers:

• The 7 custom slash skills in this repo's .claude/commands/: /pr, /review, /tdd, /test, /five, /ux, /todo
• The 2 built-in skills (/keybindings-help, /mermaid)
• Slash skills vs Agent Skills — when to use each, frontmatter contract, conversion path
• Workflow recipes — feature dev with TDD + PR, bug investigation, UX-first dev
• How to write your own skills (file format, scope, examples)
• Skills FAQ, troubleshooting, and best practices

The community has built thousands of Agent Skills. Three places to start browsing:

Notable community skills: skill-creator, skill-installer, mcp-builder, systematic-debugging, pair-programming, github-code-review, pptx, react, frontend-design, prompt-engineering-patterns, superpowers, brainstorming, market-research-reports, senior-data-engineer, and many more — see the full ecosystem section in docs/skills.md for categorized tables and install paths.

Mental model: Hooks are programmable checkpoints on Claude Code's lifecycle (before/after a tool call, session start, prompt submit, etc.). Your script inspects the proposed action and returns allow / deny / modify.

Three cases that win most teams over:

If none of those resonate, skip ahead.

Hooks live in settings files at four scopes (later overrides earlier):

Quickest setup — use the interactive menu added in 2026:

/hooks    # browse, enable, configure hooks without touching JSON

Manual setup — for the hook scripts in this repo:

1. Copy .claude/hooks/ into your project's .claude/ folder.
2. Delete the hook scripts you don't need; keep the rest.
3. Install uv (required to run the Python hook scripts).
4. Copy .claude/settings.json into your project's .claude/ folder.
5. In settings.json, replace any hardcoded uv path with the output of $(which uv).

project-root/
└── .claude/
    ├── hooks/
    │   ├── notification.py
    │   ├── post_tool_use.py
    │   └── ...
    └── settings.json

Hooks run in response to various events within Claude Code's lifecycle: examples

• PreToolUse: Runs after Claude creates tool parameters but before processing the tool call.
• PostToolUse: Runs immediately after a tool completes successfully.
• Notification: Runs when Claude Code sends notifications, such as when permission is needed to use a tool or when prompt input has been idle.
• UserPromptSubmit: Runs when the user submits a prompt, before Claude processes it.
• Stop: Runs when the main Claude Code agent has finished responding (does not run if stopped by user interrupt).
• SubagentStop: Runs when a Claude Code subagent (Task tool call) has finished responding.
• SessionEnd: Runs when a Claude Code session ends.
• PreCompact: Runs before Claude Code is about to run a compact operation.
• SessionStart: Runs when Claude Code starts a new session or resumes an existing session.
• TeammateIdle: Runs when an agent teammate becomes idle (new 2026, for Agent Teams).
• TaskCompleted: Runs when a task is marked as completed (new 2026).

Hooks receive JSON via stdin. Every event includes session_id, transcript_path, and cwd. Event-specific fields:

Two ways to communicate back: exit codes for simple control, JSON in stdout for fine-grained behavior.

Advanced: structured JSON in stdout. Per-event decision fields:

Hooks run arbitrary shell commands automatically with your user permissions — they can read, modify, or delete any file you can. Anthropic provides no warranty for what your hooks do.

Best practices:

• Validate and sanitize all inputs from stdin JSON
• Quote shell variables ("$var", not $var)
• Block path traversal (.., absolute paths outside the project)
• Use absolute paths for invoked scripts so PATH attacks don't redirect
• Explicitly skip sensitive files (.env, .git/, secrets/)

Claude Code snapshots your hook configuration at session start and warns if hooks change mid-session — review before applying.

• Timeout — 60s per hook by default; configurable per command.
• Parallelization — all matching hooks run in parallel.
• Environment — hooks run in the current dir with Claude Code's env; CLAUDE_PROJECT_DIR is available.
• Debug — /hooks shows current config; claude --debug shows hook execution logs; test scripts manually with the JSON payload piped to stdin.

Three building blocks for going beyond a single Claude session:

Git worktrees let one repo have multiple branches checked out at the same time, each in its own folder. Pair them with one Claude Code session per worktree to run independent streams of work.

git worktree add -b feature-a ../feature-a    # create the worktree
cd ../feature-a && claude                     # start Claude in it
# Repeat in another terminal for feature-b. Each session is independent.
git worktree remove ../feature-a              # clean up when done

💡 Use tmux to keep each worktree's session attached even when you close the terminal.

From your main session, ask Claude to spawn subagents for a parallel sub-task. Each subagent runs in its own context window and reports a summary back, so the main session stays focused.

Analyze the implementation of the payment feature.
Spawn 5 subagents to accelerate the work.
Ultrathink.

Specialized agents are pre-written role prompts you drop into .claude/agents/. Each one carries its own focus and tooling, so a security-reviewer agent stays focused on threats instead of also opining on code style.

This repo ships 10 production-ready specialist prompts you can drop into .claude/agents/:

📐 The two .canvas flowcharts in specialized-agents/ (agent-creation-workflow.canvas, agent-orchestration-workflow.canvas) visualize the full pipelines. Open in Obsidian or any canvas-compatible viewer.

Once you have specialized agents, the main session orchestrates them:

Have backend-engineer suggest UI improvements; have frontend-engineer
implement them; have code-reviewer review the changes; have
frontend-engineer address the review feedback.

Agent Teams is an experimental feature that lets a single Claude Code session coordinate multiple specialist agents through a shared task list. The main session acts as the team lead; teammates work on their tasks (sometimes in parallel), report progress, and update the shared list. Reach for it on full-stack features, large refactors, or anything where multiple perspectives genuinely help. Skip it for single-file edits and quick fixes.

export CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1   # add to ~/.zshrc to persist
claude

• Team lead (your main session) — assigns tasks, sets dependencies, reviews completed work.
• Teammates (specialist sub-agents) — focus on a single task each, update the shared list, can request help.
• Shared task list — single source of truth visible to everyone, tracks dependencies and status (pending / in progress / completed).

1. Parallel development — three independent streams, three teammates, one team lead synthesizing.

Task 1: frontend-engineer  – Build login UI
Task 2: backend-engineer   – Implement auth API
Task 3: qa-engineer        – Write integration tests

2. Sequential pipeline — one teammate's output feeds the next.

Task 1: backend-engineer   – Create database schema
  ↓ (blocks Task 2)
Task 2: backend-engineer   – Implement CRUD endpoints
  ↓ (blocks Task 3)
Task 3: frontend-engineer  – Build admin dashboard

3. Code review workflow — feature lands, multiple reviewers, then revision.

Task 1: feature-developer  – Implement new feature
  ↓ (completed)
Task 2: security-reviewer  – Check for vulnerabilities
Task 3: perf-reviewer      – Analyze optimization opportunities
  ↓ (both complete)
Task 4: feature-developer  – Address review feedback

4. Research → implementation — one analyst, multiple workers applying findings.

Task 1: research-agent     – Analyze existing codebase patterns
  ↓ (generates recommendations)
Task 2: implementation team – Apply findings across modules
  - Teammate A: auth module
  - Teammate B: payment module
  - Teammate C: notifications module

Team lead: coordinate overall strategy.

Teammates:
- planner: requirements → technical spec
- backend-dev: API endpoints
- frontend-dev: UI components
- db-specialist: schema design + migration
- qa-engineer: unit + integration tests
- security-reviewer: security audit

Dependencies:
1. planner → spec (blocks all)
2. db-specialist → schema (blocks backend-dev)
3. backend-dev + frontend-dev in parallel
4. qa-engineer waits for implementation
5. security-reviewer waits for all code

• Multiple agents consume more tokens — budget accordingly.
• Large teams add coordination overhead; 3–5 teammates is the sweet spot.
• Teams are session-scoped — they don't persist across restarts.
• Top performance with Opus 4.7; works on Sonnet 4.6 with shallower reasoning.

🔮 Roadmap: persistence, visual dashboards, inter-team communication. See docs/reference/changelog.md → Coming soon.

Mental model: MCP is a universal translator that lets any AI tool talk to any data source through one open protocol — USB-C for AI integrations.

Full setup walkthroughs in mcp-servers/. New to MCP? mcp-servers/playwright.md has a working three-line setup.

See mcp-servers/README.md for the comparison matrix, install commands, and troubleshooting.

Install from the registry or follow the source link. No walkthrough yet — source READMEs cover setup.

General-purpose:

Specialized — reach for these when the workflow fits:

Before MCP, every AI app needed a custom integration for every tool: n apps × m tools = n × m brittle one-off connections. Teams inside the same company would reinvent the same Slack/GitHub/Postgres integration over and over.

MCP collapses this to N + M: each app implements MCP once, each tool exposes MCP once, and any combination works together. Same pattern Web APIs gave us for app-to-server and LSP gave us for editor-to-language tooling.

Each pillar makes ownership explicit, so it's always clear who's driving:

The official MCP Registry (live since September 2025) is the app-store-equivalent for MCP servers. An agent that needs to check Grafana logs but doesn't have a Grafana tool wired up can ping the registry, find the verified server, install it, and continue — teaching itself a new capability on the fly.

• Live registry — search official and community servers at registry.modelcontextprotocol.io or claude mcp search <keyword>.
• Linux Foundation governance — MCP donated to the Agentic AI Foundation for vendor-neutral development.
• MCP Apps — servers can ship interactive UI components, not just text tools.
• OAuth client credentials — built-in --client-id / --client-secret flags for authenticated services.
• Async operations — non-blocking server calls for long-running tasks.
• Server discovery — .well-known URLs for automatic discovery.

claude mcp add <server> npx '@<package>@latest'                    # install from registry
claude mcp add my-api --client-id ID --client-secret S npx '...'   # with OAuth
/mcp                                                                # list connected servers

📚 References: MCP roadmap · Official servers.

/fast toggles 2.5× faster responses at 6× the price. The ↯ indicator confirms it's on.

⚠️ Fast Mode runs on Opus 4.6 — not 4.7. Even if your default model is Opus 4.7, /fast switches the underlying model to 4.6. Sonnet and Haiku don't have a Fast Mode.

/fast                                    # toggle on (↯ appears)
> fix the auth bug in src/login.ts       # ~2.5× faster
/fast                                    # toggle off when done

Decision rule: use it when latency matters (live debugging, demo prep, time-pressured fixes). Skip it for background work, exploration, or anything where 2.5× faster isn't worth 6× the cost. Use /cost to monitor the bill.

An open-source framework that bolts pre-built personas, commands, and workflows on top of Claude Code.

What it is. Super Claude ships 15+ ready-made specialist agents (architect, security, performance, frontend, etc.) plus a library of structured slash commands (/sc:analyze, /sc:improve, /sc:design) and behavioral flags. It installs into ~/.claude/ as a drop-in extension to Claude Code.

Why pair it with Claude Code. Skips the "write your own role prompts" phase. You get a curated library of expert personas and commands the community has already iterated on — useful as both a daily tool and a reference for how to write your own skills.

When to reach for it. You want pre-built specialist agents without building them yourself, or you're learning prompt patterns by reading well-crafted examples.

→ Source on GitHub

A multi-agent framework that walks a project from idea to working software using a team of Claude agents in defined roles.

What it is. BMAD (Breakthrough Method of Agile AI-driven Development) coordinates specialist agents — analyst, PM, architect, developer, QA, scrum master — through a full SDLC. Each role produces specific artifacts (PRD, architecture doc, story file) that the next role consumes. Works as a Claude Code extension via skills and agents.

Why pair it with Claude Code. Gives you a documented, repeatable workflow for greenfield builds. Instead of asking Claude "build me an app," you walk through analyst → PM → architect → developer → QA, each agent producing structured output the next one consumes.

When to reach for it. Greenfield projects, large feature builds, or any time you want planning rigor before code. Overkill for a quick fix or a one-file refactor.

→ Source on GitHub

→ Full FAQ in docs/reference/faq.md — covers models, pricing, tokens, plans, Fast Mode, worktrees, and Pro-plan optimization.

A few of the most-asked questions:

How many messages do I get on the Pro plan? ~45 messages per 5-hour rolling window. Full access to all models (Opus 4.7, Sonnet 4.6, Haiku 4.5; previous-gen 4.6/4.5 also available). Details →

What's the difference between Pro, Max 5x, and Max 20x? Pro $20/mo, Max 5x $100/mo (5× usage), Max 20x $200/mo (20× usage). All tiers include Claude Code and all models. Pricing details →

Should I use Fast Mode? Use it for rapid iteration and time-sensitive work; skip it for long background tasks (6× pricing). More →

What's the difference between custom slash commands and skills? Same thing — both are markdown files in .claude/commands/. See Skills FAQ in docs/skills.md.

Can I use Opus 4.7's 1M context window? Beta and API-only at the moment. Subscription plans are limited to 200K. Beta access details →

→ Full changelog in docs/reference/changelog.md — major changes, new features, and deprecations through May 2026.

Recent highlights:

• 🆕 Claude Opus 4.7 (May 2026) — sharper reasoning and better long-context grounding than Opus 4.6 at the same price.
• 🆕 Claude Sonnet 4.6 (May 2026) — replaces Sonnet 4.5 as the everyday balanced tier.
• 🆕 Agent Teams — multi-agent collaboration with team leads and shared task lists (experimental).
• 🆕 MCP Registry is live at registry.modelcontextprotocol.io, now governed by the Linux Foundation.
• 🆕 New commands: /fast, /auth, /debug, /teleport, /rename, /hooks.
• 📝 Pro plan now includes every current and previous-generation model; usage measured in messages (~45 / 5hr), not tokens.
• ❌ /login and /logout deprecated in favor of /auth login / /auth logout.

💡 For Anthropic's authoritative release notes, see docs.anthropic.com/release-notes/claude-code.

• Reading on GitHub? The in-page anchor links work natively. Tap the table-of-contents button (top-left of the file view) for fast jumping.
• Want a richer markdown view? This repo plays well with Obsidian — handy if you also want to open the specialized-agents/*.canvas flowcharts.
• On a phone? Stick to the Choose your path section at the top; the wider tables read best on desktop.

→ Full reading list in docs/reference/further-reading.md

A curated set of pointers — official Anthropic docs, MCP resources, hooks examples, workflow tutorials, pricing references, and adjacent tooling.

Most-clicked starting points:

• Claude Code overview (official)
• Claude Code best practices (Anthropic engineering)
• Building effective agents (Anthropic engineering)
• Official MCP Registry
• Hooks reference (official)

Features, pricing, and availability change frequently. Always check the official Anthropic documentation for the most current information.

Last reviewed May 2026 · Spotted something stale? Open an issue or send a PR — see CONTRIBUTING.md.