New AI Tools
banner

codemcp


Introduction:

codemcp is a programming assistant for Claude Pro, capable of directly editing code, running tests, eliminating the need for code copying, and ensuring modifications are rollback-able through Git.









codemcp

codemcp is a tool designed to turn Anthropic's Claude Desktop application into a more efficient pair programming assistant. It allows Claude to implement features, fix bugs, and perform refactoring directly in your local codebase without the need to frequently copy and paste code between the chat window and the IDE.

Main Features:

  • Cost Efficiency: Relies on a Claude Pro subscription, avoiding the high costs associated with using API keys.
  • Secure Agentic AI: By providing a limited and secure set of tools and enforcing best practices like Git version control, it minimizes the risk of AI misuse, allowing users to confidently let the AI work and then evaluate and decide whether to accept the changes.
  • IDE Agnostic: Seamlessly integrates with your existing IDE, so after Claude modifies the code, you can review the changes and make further edits in your preferred IDE.

Use Cases:

  1. Feature Development: Let Claude assist you in implementing new features, such as adding new API interfaces or UI components.
  2. Bug Fixing: Guide Claude to fix defects in the code, such as program crashes or data errors.
  3. Code Refactoring: Delegate Claude to refactor code, such as extracting duplicate code or optimizing algorithm logic.

How to Use:

  1. Install uv.
  2. Configure claude_desktop_config.json, adding the codemcp server configuration.
  3. Create a codemcp.toml file in your Git repository to configure formatting and testing commands.
  4. In Claude Desktop, create a project and initialize the project directory in the project description.
  5. Communicate with Claude about the changes you want to make to the project. codemcp will generate a commit each time the code is changed.

With codemcp, users can collaborate more conveniently with Claude, improving development efficiency without worrying about excessive API costs, while ensuring the security and controllability of the code.