📋 Table of Contents

• What is cp-toolkit?
• Quick Start
• Project Types
• Features
• Commands
• Generated Structure
• Usage with GitHub Copilot
• Contributing
• Roadmap
• License

🎯 What is cp-toolkit?

cp-toolkit is a CLI tool that bootstraps AI agent configurations for your projects. It creates a standardized structure that GitHub Copilot and other AI assistants understand, making them smarter and more context-aware when working with your codebase.

Think of it as ESLint for AI — it defines rules, specialists, and behaviors that AI assistants follow when helping you code.

Why cp-toolkit?

📦 Quick Start

Installation

# Install globally
npm install -g cp-toolkit

# Or use with npx (no install needed)
npx cp-toolkit init

Initialize your project

cd your-project
cp-toolkit init

You'll be prompted to select your project type and name. That's it! Your project now has AI-powered agent configurations.

🏗️ Project Types

When running cp-toolkit init, you'll be asked to choose a project type. Each type optimizes the agent configuration for your use case:

Single App

Standard single application project.

• Best for: Websites, web apps, standalone applications
• Structure: Agents configured at project root
• Examples: Next.js app, React SPA, Express server

Monorepo

Multi-package repository with shared configurations.

• Best for: Turborepo, Nx, Lerna, npm/pnpm workspaces
• Structure: Root-level agents + per-package customization possible
• Examples: Design system + docs + app, microservices

Library/Package

Publishable package or library.

• Best for: npm packages, SDKs, frameworks, component libraries
• Focus: Documentation, testing, API design, compatibility
• Examples: UI component library, utility package, SDK

API Only

Backend-only project without frontend.

• Best for: REST APIs, GraphQL servers, microservices
• Focus: Backend, database, security, and DevOps agents
• Examples: Express API, FastAPI, NestJS backend

✨ Features

🤖 20+ Specialized Agents

Each agent is an expert in a specific domain:

📋 Path-Specific Instructions

Instructions are automatically applied based on file patterns:

# .github/instructions/typescript.instructions.md
---
applyTo: "**/*.ts,**/*.tsx"
---

## TypeScript Guidelines
- Enable strict mode
- No `any` types
- Use discriminated unions

🔌 MCP Integration

Out-of-the-box Model Context Protocol support for VS Code:

// .vscode/mcp.json (auto-generated)
{
  "servers": {
    "filesystem": { ... },
    "memory": { ... },
    "sequentialThinking": { ... }
  }
}

🛠️ Commands

cp-toolkit init [directory]

Initialize cp-toolkit in a project.

cp-toolkit init              # Current directory (interactive)
cp-toolkit init my-project   # New directory
cp-toolkit init -y           # Skip prompts, use defaults
cp-toolkit init -f           # Force overwrite existing config

Options:

cp-toolkit add <type> <name>

Add new components to your project.

cp-toolkit add agent rust-specialist     # Add custom agent
cp-toolkit add instruction golang        # Add language instruction

Types:

• agent - Create a new specialized agent
• instruction - Create path-specific instructions

cp-toolkit list [type]

List available components.

cp-toolkit list              # List everything
cp-toolkit list agents       # List all agents
cp-toolkit list instructions # List all instructions

cp-toolkit doctor

Diagnose your configuration and check for issues.

cp-toolkit doctor

Example output:

✓ .github/ directory exists
✓ copilot-instructions.md exists
✓ 20 agents found
✓ 5 instructions found
✓ AGENTS.md exists at root
✓ .vscode/mcp.json exists
✨ cp-toolkit is healthy!

📂 Generated Structure

After running cp-toolkit init, your project will have:

your-project/
├── .github/
│   ├── copilot-instructions.md    # Global AI instructions (always active)
│   ├── agents/                    # Agent definitions
│   │   ├── orchestrator.md
│   │   ├── frontend-specialist.md
│   │   ├── backend-specialist.md
│   │   ├── database-architect.md
│   │   ├── security-auditor.md
│   │   └── ... (20+ agents)
│   └── instructions/              # Path-specific rules
│       ├── typescript.instructions.md
│       ├── python.instructions.md
│       ├── database.instructions.md
│       └── security.instructions.md
├── .vscode/
│   └── mcp.json                   # MCP server configuration
└── AGENTS.md                      # Universal AI instructions

🎮 Usage with GitHub Copilot

After initialization, invoke agents in GitHub Copilot Chat using @agent-name:

@frontend-specialist Create a responsive navbar with dark mode toggle

@security-auditor Review this authentication middleware for vulnerabilities

@orchestrator Implement a user dashboard with profile settings and activity feed

@database-architect Design a schema for a multi-tenant SaaS application

@test-engineer Write comprehensive tests for the UserService class

Tips

• Use @orchestrator for complex, multi-domain tasks
• Combine agents: ask @security-auditor to review code written by @backend-specialist
• Path-specific instructions activate automatically based on the file you're editing

🤝 Contributing

We'd love your help making cp-toolkit better!

🐛 Report Bugs

Found a bug? Open an issue with:

• Your OS and Node.js version
• Steps to reproduce
• Expected vs actual behavior

💡 Suggest Features

Have an idea? Start a discussion or open an issue.

🔧 Submit PRs

1. Fork the repository
2. Create your feature branch: git checkout -b feature/amazing-feature
3. Commit your changes: git commit -m 'feat: add amazing feature'
4. Push to the branch: git push origin feature/amazing-feature
5. Open a Pull Request

🤖 Add New Agents

Create specialized agents for new domains:

cp-toolkit add agent rust-specialist
# Then edit .github/agents/rust-specialist.md

Share your agents by submitting a PR!

🌍 Add Language Instructions

Help developers with new language support:

cp-toolkit add instruction golang
cp-toolkit add instruction kotlin
cp-toolkit add instruction swift

🗺️ Roadmap

• cp-toolkit sync — Sync with remote agent repositories
• cp-toolkit upgrade — Update agents to latest versions
• Plugin system for custom MCP servers
• VS Code extension for visual management
• AI-powered agent suggestions based on codebase analysis
• Team sharing via private registries

📜 License

MIT © 2026 gensart-projs