CodeCraft CLI 🚀

AI Governance & Assistant CLI for Developers Manage your project's architecture, planning, and code quality with an AI-powered CLI.

CodeCraft is a next-generation CLI that brings LLM-powered reasoning to your terminal. It acts as a Project Manager, Architect, and QA Lead, helping you build software faster and with higher quality.

✨ Features

• Interactive REPL: A "Claude Code" style interactive shell with persistent context.
• Role-Based AI: Switch between personas (PM, Architect, Backend/Frontend Developer, QA).
• Slash Commands: Trigger workflows effortlessly (/plan, /code, /verify, /doctor).
• Scope Guard: Prevents AI from modifying files outside the allowed scope.
• Rule Engine: Enforces project-specific architectural rules.

📦 Installation

npm install -g @gkganesh12/codecraft-cli

🚀 Quick Start

1. Initialize CodeCraft in your project:

   codecraft init

2. Start the Interactive Mode:

   codecraft

   You will see the CodeCraft > prompt.

3. Try a Command:

   /plan "Create a user authentication API with JWT"
   

🛠️ Usage

Interactive Commands (REPL)

Command	Description
/init	Initialize CodeCraft in the current directory.
/plan <task>	Create a detailed implementation plan using the Project Manager persona.
/architect	Review the current plan against architecture rules using the Architect persona.
/code	Generate code based on the approved plan using Developer personas.
/verify	Analyze risks and suggest tests using the QA persona.
/feature <desc>	Draft a comprehensive Feature Specification.
/adr <title>	Create an Architecture Decision Record (ADR).
/test <file>	Generate unit tests for a specific file.
/review <file>	Perform an AI code review on a file.
/explain <file>	Explain complex code in plain English.
/config	Manage CLI settings (API keys, models, verbose mode).
/doctor	Check environment health and dependencies.
/update	Check for the latest version of CodeCraft.

CLI Flags

You can also run commands directly without entering the REPL:

# Run a specific command
codecraft doctor

# Run with verbose logging
codecraft --verbose plan "Refactor the database schema"

# Override the AI model for a session
codecraft --model gpt-4o code

⚙️ Configuration

CodeCraft stores its configuration in ~/.codecraft/config.json. You can manage it via:

codecraft config

Supported settings:

• API Key: Your OpenAI or supported LLM provider key.
• Model: Default model to use (e.g., gpt-4-turbo, gpt-4o).
• Verbose: Enable detailed debug logs.

🤝 Contributing

Contributions are welcome! Please look at the Docs/ folder for architectural guidelines.

1. Fork the repository.
2. Create a feature branch.
3. Commit your changes.
4. Open a Pull Request.

📄 License

MIT

Created with ❤️ By Ganesh Khetawat for Developers

Feel Free to Tweak and make changes

Request a Feature