KiCad MCP Server
Enables natural language search and exploration of KiCad component symbol libraries with fast full-text search across 20,000+ components including metadata like datasheets, footprints, and descriptions.
README
KiCad MCP Server
A Model Context Protocol (MCP) server that provides searchable access to KiCad component symbol libraries. Query your local KiCad libraries using natural language through Claude or other MCP-compatible clients.
Features
- Fast full-text search across 20,000+ KiCad components using SQLite FTS5
- Component metadata including references, values, descriptions, datasheets, and footprints
- Separate indexer and server for optimal performance
- MCP-compliant tools for integration with Claude Code and other MCP clients
Installation
# Clone the repository
git clone https://github.com/yourusername/kicad_mcp.git
cd kicad_mcp
# Install using pip
pip install -e .
# Or install with development dependencies
pip install -e ".[dev]"
Quick Start
1. Index Your KiCad Libraries
First, build the component database from your KiCad symbol libraries:
# Index default KiCad locations (/usr/share/kicad/symbols/)
kicad-index
# Or specify a custom directory
kicad-index /path/to/your/symbols
# Check database statistics
kicad-index --stats
The database is stored at ~/.local/share/kicad-mcp/components.db.
2. Configure MCP Server
For Claude Code
Add the server to your Claude Code configuration:
claude mcp add --transport stdio kicad --scope user -- /path/to/kicad_mcp/.venv/bin/kicad-mcp
Or manually edit ~/.claude/claude_mcp_config.json:
{
"mcpServers": {
"kicad": {
"command": "/path/to/kicad_mcp/.venv/bin/kicad-mcp",
"args": []
}
}
}
For Other MCP Clients
The server communicates via stdio using the MCP protocol. Configure your client to launch kicad-mcp as a subprocess.
3. Use the Tools
Once configured, you'll have access to these MCP tools:
search_components- Search by name, description, or keywordslist_component_types- Get all reference designator types (R, C, U, etc.)get_components_by_type- Filter components by reference typeget_component_details- Get complete metadata for a specific component
Examples
Searching for Components
Find me an ATmega microcontroller
-> Returns ATmega328, ATmega32U4, etc. with datasheets and footprints
What capacitors are available?
-> Lists capacitors with descriptions and package options
I need a voltage regulator
-> Shows LDOs, switching regulators, etc.
Component Details
Each component includes:
- Name - Component identifier
- Library - Source KiCad library
- Reference - Designator prefix (R, C, U, IC, etc.)
- Value - Component value/model
- Description - Human-readable description
- Keywords - Searchable tags
- Datasheet - URL to datasheet (if available)
- Footprint - Associated PCB footprint
Architecture
This project uses a two-tool architecture:
- Indexer (
kicad-index) - Parses.kicad_symfiles and builds a SQLite database with full-text search - MCP Server (
kicad-mcp) - Long-running process that serves queries from the pre-built database
This separation ensures fast server startup and allows updating the index independently.
KiCad Library Locations
Default search paths (Linux):
- System libraries:
/usr/share/kicad/symbols/ - User libraries:
~/.local/share/kicad/9.0/symbols/
Adjust paths for your OS and KiCad version.
Development
# Create virtual environment (using uv)
uv venv
source .venv/bin/activate
# Install in development mode
uv pip install -e ".[dev]"
# Run tests
pytest
# Rebuild index after making changes
kicad-index
Requirements
- Python 3.11+
- KiCad symbol libraries (
.kicad_symfiles) - Dependencies:
sexpdata,mcp(installed automatically)
How It Works
- Parsing: The indexer uses
sexpdatato parse KiCad's s-expression format - Storage: Components are stored in SQLite with FTS5 full-text search indexes
- Querying: The MCP server receives requests via stdio and queries the database
- Response: Results are returned in MCP-compliant JSON format
License
MIT
Contributing
Contributions welcome! Please open an issue or pull request.
Acknowledgments
- Built using Anthropic's MCP SDK
- Parses KiCad symbol library format
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。