KiCad MCP Server

This project is a Model Context Protocol (MCP) server for KiCad. As MCP server that utilizes KiCad's official IPC-API, it provides the most stable and reliable way for AI models like Claude to interact with KiCad, automating and assisting with PCB design and schematic tasks.

Key Features

• MCP Server Implementation: Handles requests from MCP clients.
• KiCad Integration: Communicates with a running KiCad session using the kicad-python library.
• Automated Workflows: Enables AI models to create, modify, and verify schematics and PCB layouts in KiCad projects.

Wiring

https://github.com/user-attachments/assets/e2ba57e7-2c77-4c56-a911-c461c77307e4

Move item

https://github.com/user-attachments/assets/de6c93dc-8808-4321-827e-ebad0556e7b1

Enhanced Board Analysis and Action Verification

https://github.com/user-attachments/assets/0fea60de-d012-4b4d-bfa4-dd1b758b2c7f

This server features AnalyzeTools for enhanced board analysis. A key tool is get_board_status, which provides a comprehensive overview of the PCB layout, including screenshots.

To ensure robust and reliable operations, the server uses an ActionFlowManager. This manager orchestrates the execution of actions by automatically invoking get_board_status both before and after each action. This flow provides the AI model with critical context:

• Pre-Action Analysis: By reviewing the board's state before an action, the model can make better-informed decisions.
• Post-Action Verification: By comparing the board's state before and after the action, the model can visually confirm that the operation was successful and achieved the desired outcome.

This automated verification process significantly improves the accuracy and reliability of automated PCB design tasks.

What Can It Do?

With this MCP server, an AI model can perform tasks such as:

• Manipulate PCB Objects:
  • Create new items (footprints, tracks, etc.)
  • Modify the properties of existing items
  • Move and rotate items
  • Delete unnecessary items
• Analyze PCBs:
  • Get a list of all items of a specific type on the board.
  • Query the overall status information of the board using get_board_status for comprehensive analysis.

Getting Started

This project uses Poetry to manage dependencies.

1. Prerequisite: Install kicad-python

This project uses the kicad-python library as a Git submodule. Therefore, you must build and install kicad-python before running this project.

1. Clone Repository and Initialize Submodules: Run git submodule update --init to add kicad-python's source code as a submodule.

   git submodule update --init --depth 1
   

2. Build and Install kicad-python: Navigate to the kicad-python directory and follow the instructions in that project's COMPILING.md file to build and install the library.

2. Configure Environment Variables

Before running the server, you need to create a .env file in the project's root directory (KiCad-mcp-python/.env). This file is crucial for tools that rely on KiCad's command-line interface (CLI), such as get_board_status which generates screenshots to provide visual context of the board. It stores the necessary environment variables for the server to function correctly.

Create a file named .env and add the following content, adjusting the paths to match your system configuration:

KICAD_CLI_PATH=/path/to/your/kicad-cli
PCB_PATHS=/path/to/your/project1.kicad_pcb,/path/to/your/project2.kicad_pcb

Variable Explanations:

• KICAD_CLI_PATH: The absolute path to the KiCad command-line interface (CLI) executable.
  • macOS Example: /Applications/KiCad/KiCad.app/Contents/MacOS/kicad-cli
  • **Windows Exam

... View full README on GitHub