codebuddy-cli

What This Skill Does

The CodeBuddy CLI skill provides an interface to Tencent's AI-powered programming assistant directly within your terminal. It leverages advanced language models to perform complex coding tasks, refactor existing logic, and debug errors through natural language prompts. By integrating seamlessly into your development workflow, it eliminates the need to context-switch between your code editor and a web-based AI interface. Whether you are generating boilerplate code, writing comprehensive unit tests, or auditing your project's security, this tool acts as an autonomous pair programmer residing in your shell.

Installation

To begin using CodeBuddy, ensure you have Node.js version 18 or higher installed. The installation is performed globally via npm, which registers the codebuddy command system-wide:

1. Verify your Node.js environment: node -v.
2. Install the package using: npm install -g @tencent-ai/codebuddy-code.
3. Validate the installation by running codebuddy --version in your terminal.

Once installed, you can initialize a session by navigating to your project directory and typing codebuddy. You will be prompted to authenticate via Google/GitHub for international models or WeChat for China-specific DeepSeek models.

Use Cases

CodeBuddy is designed for developers who need on-the-fly assistance without leaving the CLI. Key use cases include:

• Automated Refactoring: Quickly optimize functions or improve code readability by providing a brief natural language request.
• Test Generation: Generate robust unit tests for new API endpoints or refactored legacy code modules.
• Project Analysis: Use the /init command to generate automated project documentation or /doctor to diagnose workspace-specific environment issues.
• CI/CD Integration: Streamline code quality checks before pushing commits by running inspection tasks directly from the terminal.

Example Prompts

1. "Generate a suite of Jest unit tests for the user authentication service located in src/auth.ts."
2. "Review the current directory for common security vulnerabilities and suggest a mitigation plan."
3. "Refactor the database connection logic to use a connection pool instead of individual instances."

Tips & Limitations

• Security: The --dangerously-skip-permissions flag is provided for sandbox environments only. Never use this in production, as it permits the AI to perform file operations without your approval, which may lead to data loss or unwanted file deletion.
• Memory: Utilize the /memory command to provide the AI with context about your project architecture to improve response accuracy.
• Customization: Extend the functionality by adding custom scripts to the .codebuddy/commands/ directory in your project folder, allowing for project-specific automation that persists across sessions.