Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
Claude Code plugin pack for commands, roles, and hooks
This repository packages Claude Code commands, expert roles, and hook scripts into an installable plugin system. It is meant to reduce repetitive setup and let Claude Code handle tasks like PR creation, code review, refactoring, dependency updates, and status checks.
Builders who use Claude Code and want a shared set of commands, roles, and workflow scripts.
You can handle routine coding and PR work with the same reusable Claude Code setup instead of repeating the same instructions each time.
What it does
Slash commands
39 commands for PR management, code quality, refactoring, planning, dependency updates, and error fixing.
Role presets
Named roles like security, architect, frontend, backend, performance, qa, mobile, reviewer, and analyzer that can run in the main context or as sub-agents.
Hooks
Event-triggered scripts in `settings.json`, including file-permission preservation around tool use.
Language plugins
Localized plugin folders for English, Japanese, Korean, Simplified Chinese, Traditional Chinese, Spanish, French, and Portuguese.
Marketplace packaging
A Claude Code marketplace entry that lets you add and install the cookbook as a plugin.
How to get it
- 1First, add this repository as a plugin marketplace in Claude Code
/plugin marketplace add wasabeef/claude-code-cookbook
README
Claude Code Cookbook
A collection of commands, roles, and automation scripts for Claude Code.
Automate your workflow without unnecessary confirmations, allowing you to focus on what matters. Claude Code judges and executes common tasks like code fixes, test runs, and documentation updates.
Read in your language: 🇯🇵 日本語 · 🇺🇸 English · 🇰🇷 한국어 · 🇨🇳 简体中文 · 🇹🇼 繁體中文 · 🇪🇸 Español · 🇫🇷 Français · 🇧🇷 Português
What is Claude Code Cookbook?
Claude Code Cookbook provides a plugin system that extends Claude Code with:
- Commands: Custom slash commands for common development tasks
- Roles: Expert role presets for specialized assistance
- Hooks: Automated scripts that trigger at specific events
Key Features
Commands
39 slash commands organized by category. Execute by typing / followed by the command name.
Pull Request Management
| Command | Description |
|---|---|
/pr-list | Display prioritized list of open PRs in current repository |
/pr-issue | Display prioritized list of open Issues in current repository |
/pr-create | Auto-generate PR from Git changes with detailed description and optimal branch setup |
/pr-review | Systematic code quality and architecture review for Pull Requests |
/pr-fix | Efficiently respond to PR review comments with 3-stage error analysis |
/pr-auto-update | Automatically update PR description and labels based on changes |
Code Quality & Refactoring
| Command | Description |
|---|---|
/refactor | Safe, step-by-step code refactoring with SOLID principles evaluation |
/smart-review | Advanced code review to improve quality |
/tech-debt | Analyze technical debt and create prioritized improvement plans |
/analyze-dependencies | Analyze project dependencies and visualize circular dependencies |
/analyze-performance | Analyze application performance issues and propose improvements |
/design-patterns | Propose and review implementations based on design patterns |
Development Tools
| Command | Description |
|---|---|
/fix-error | Suggest code fixes based on error messages |
/explain-code | Clearly explain functionality and logic of selected code |
/commit-message | Generate commit messages based on changes |
/semantic-commit | Split large changes into meaningful units with semantic commit messages |
/pr-checks | Monitor GitHub Actions CI status and track until completion |
/screenshot | Capture and analyze screen screenshots |
Planning & Analysis
| Command | Description |
|---|---|
/plan | Activate planning mode and formulate detailed implementation strategies |
/spec | Create detailed specifications from requirements (spec-driven development) |
/ultrathink | Execute structured thinking process for complex issues |
/check-fact | Verify information accuracy by referencing codebase and documentation |
/sequential-thinking | Think through complex problems step-by-step using Sequential Thinking MCP |
/task | Launch specialized agents for autonomous search, research, and analysis tasks |
Dependency Management
| Command | Description |
|---|---|
/update-node-deps | Safely update dependencies in Node.js projects |
/update-flutter-deps | Safely update dependencies in Flutter projects |
/update-rust-deps | Safely update dependencies in Rust projects |
See your language plugin for the complete list of 39 commands with detailed documentation.
Roles
Switch Claude to expert roles for specialized assistance. Each role can run independently as a sub-agent using the --agent option for parallel execution without interfering with main context.
| Role | Description |
|---|---|
security | Security vulnerability analysis and threat detection |
architect | Software architecture design and system design patterns |
frontend | UI/UX optimization and frontend best practices |
backend | API design, microservices, and cloud-native architecture |
performance | Performance optimization for speed and memory |
qa | Test planning and quality assurance strategies |
mobile | iOS/Android development and mobile-first design |
reviewer | Code review focusing on readability and maintainability |
analyzer | System analysis and root cause analysis |
Usage Examples
# Normal mode (execute in main context)
/cook:role security
"Review this authentication system for vulnerabilities"
# Sub-agent mode (execute in independent context)
/cook:role security --agent
"Perform comprehensive security audit of entire project"
# Multiple roles in parallel
/cook:multi-role security,performance --agent
"Analyze system's security and performance comprehensively"
Hooks
Automate development workflow with event-triggered scripts configured in settings.json:
| Hook Script | Event | Description |
|---|---|---|
preserve-file-permissions.sh | PreToolUse / PostToolUse | Save and restore file permissions to prevent unintended changes |
Development Workflow
Typical development flow using Claude Code Cookbook commands:
flowchart TB
Start([Task Confirmation]) --> PRList["/cook:pr-list<br/>Open PR List"]
Start --> PRIssue["/cook:pr-issue<br/>Open Issue List"]
PRList --> TaskType{Type?}
PRIssue --> TaskType
TaskType -->|New Feature| Plan["/cook:spec<br/>Requirements & Design"]
TaskType -->|Bug Fix| Fix["/cook:fix-error<br/>Error Analysis"]
TaskType -->|Refactoring| Refactor["/cook:refactor<br/>Improvement"]
TaskType -->|Review| Review["/cook:pr-review<br/>Review"]
Plan --> Design["/cook:role architect<br/>/cook:role-debate<br/>Design Consultation"]
Design --> Implementation[Implementation & Testing]
Fix --> Implementation
Refactor --> Implementation
Review --> Implementation
Implementation --> Check["/cook:smart-review<br/>Quality Check"]
Check --> Commit["/cook:semantic-commit<br/>Commit by Purpose"]
Commit --> PR["/cook:pr-create<br/>Auto PR Creation"]
PR --> CI["/cook:pr-checks<br/>CI Status Check"]
CI --> Status{Issues?}
Status -->|Yes| Feedback["Fix Response<br/>/cook:pr-fix<br/>/cook:fix-error"]
Status -->|No| End([Complete])
Feedback --> Implementation
classDef commandBox fill:#e0f2fe,stroke:#0369a1,stroke-width:2px,rx:5,ry:5,color:#0c4a6e
classDef processBox fill:#f0f9ff,stroke:#0ea5e9,stroke-width:1px,rx:5,ry:5,color:#075985
classDef decisionBox fill:#fef3c7,stroke:#f59e0b,stroke-width:2px,rx:5,ry:5,color:#78350f
classDef startEnd fill:#86efac,stroke:#22c55e,stroke-width:2px,rx:20,ry:20,color:#14532d
class PRList,PRIssue,Plan,Fix,Refactor,Review,Design,Check,CI,Commit,PR,Feedback commandBox
class Implementation processBox
class TaskType,Status decisionBox
class Start,End startEnd
Installation
Step 1: Add the Marketplace
First, add this repository as a plugin marketplace in Claude Code:
/plugin marketplace add wasabeef/claude-code-cookbook
Step 2: Install Your Language Plugin
Choose and install your preferred language plugin:
| Language | Plugin Name | Install Command |
|---|---|---|
| 🇯🇵 日本語 | plugins/ja | /plugin install cook@claude-code-cookbook |
| 🇺🇸 English | plugins/en | /plugin install cook-en@claude-code-cookbook |
| 🇰🇷 한국어 | plugins/ko | /plugin install cook-ko@claude-code-cookbook |
| 🇨🇳 简体中文 | plugins/zh-cn | /plugin install cook-zh-cn@claude-code-cookbook |
| 🇹🇼 繁體中文 | plugins/zh-tw | /plugin install cook-zh-tw@claude-code-cookbook |
| 🇪🇸 Español | plugins/es | /plugin install cook-es@claude-code-cookbook |
| 🇫🇷 Français | plugins/fr | /plugin install cook-fr@claude-code-cookbook |
| 🇧🇷 Português | plugins/pt | /plugin install cook-pt@claude-code-cookbook |
Each plugin includes:
- Full command documentation in the native language
- Role definitions optimized for that language
- Language-specific customization examples
Quick Start
After installation, commands and agents are invoked with a plugin-specific prefix:
# Commands (example with Japanese plugin)
/cook:pr-create
/cook:semantic-commit
/cook:check-fact
# Agents
@agent-cook:frontend
/cook:role security
# Explore available commands
/cook:
Note: The plugin name is automatically added as a prefix to prevent namespace conflicts when multiple language plugins are installed.
Why Claude Code Cookbook?
- Multilingual: 8 language plugins with native translations
- Production-ready: Battle-tested commands and roles
- Comprehensive: 39 commands covering entire development workflow
- Extensible: Easy to add your own commands and roles
- Safe: Built-in safety hooks prevent dangerous operations
- Efficient: Reduce repetitive tasks and context switching
- Open Source: Apache 2.0 licensed
Example Workflow
# 1. Check current tasks
/cook:pr-list
# 2. Plan new feature
/cook:spec
/cook:role architect
# 3. Implement with quality checks
# ... write code ...
/cook:smart-review
# 4. Commit changes logically
/cook:semantic-commit
# 5. Create PR automatically
/cook:pr-create
# 6. Monitor CI and respond to feedback
/cook:pr-checks
/cook:pr-fix
Files in the repo
- .claude-plugin
- .github
- assets
- plugins
- scripts
- .gitignore
- .markdownlint.json
- .mcp.json
- commitlint.config.js
- lefthook.yml
- package-lock.json
- package.json
- README.md
- settings.json
Discussion (0)
Ask about usage, or say what you built with itSign in to join the discussion.
No comments yet. Be the first to say what this is good for.
More plugins

Graphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more.
OmO: Just type "mass ulw" keyword with your prompt. Now you are the master of graph engineering.
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
Opinionated Oxlint rules for rejecting low-evidence TypeScript and JavaScript patterns
Teams-first Multi-agent orchestration for Claude Code