mcp4eda

mcp4eda is an MCP server collection that integrates electronic design automation (EDA) tools into Model Context Protocol clients like Claude Desktop. It connects LLM-driven environments directly to standard open-source hardware engineering utilities including Yosys, Verilator, Verible, GTKWave, KLayout, and OpenLane, as well as domain calculators like AnySilicon. Digital design engineers, ASIC architects, and FPGA developers use this suite to run logic synthesis, execute C++ based RTL simulations, run syntax linting, visualize waveforms, inspect layout files, and execute automated RTL-to-GDSII pipelines. Rather than manually switching across terminal environments and graphic interfaces, users can issue natural language prompts or structured instructions to inspect register hierarchies, compute die counts per wafer, run Design Rule Checks (DRC), and query semiconductor IP vendors. By bridging the gap between conversational AI and foundational hardware compilers, mcp4eda enables rapid inspection, automated testbench generation, and continuous verification directly inside developer workflows.

Category: Developer Tools & Code Intelligence

Tags: eda, hardware-design, semiconductor, systemverilog, verilog

Visit mcp4eda

How to install and configure mcp4eda

1. Clone the repository: bash git clone https://github.com/ssql2014/mcp4eda.git cd mcp4eda 2. Install and build all server components: bash for dir in yosys-mcp verilator-mcp verible-mcp gtkwave-mcp klayout-mcp anysilicon semiconductor-supply-chain-mcp openlane-mcp; do cd $dir && npm install && npm run build && cd .. done 3. Ensure prerequisites such as Node.js 18+ and relevant EDA binaries (Yosys, Verilator, Verible, GTKWave, KLayout) are installed on your system. 4. Open your Claude Desktop configuration file (claude_desktop_config.json) and add the server definitions under mcpServers. For example: json { "mcpServers": { "yosys": { "command": "node", "args": ["/path/to/mcp4eda/yosys-mcp/dist/index.js"], "env": { "YOSYS_PATH": "/opt/homebrew/bin/yosys", "LOG_LEVEL": "info" } } } } 5. Save the file and restart Claude Desktop.

What you can do with mcp4eda

• Synthesizing Verilog and SystemVerilog designs to gate-level netlists using Yosys targeting Xilinx, Intel, or Lattice FPGAs. * Simulating RTL code with Verilator to run automated testbench generation and parse execution coverage. * Linting and formatting SystemVerilog code with Verible to enforce consistent styling and inspect abstract syntax trees. * Executing Design Rule Checking (DRC) and extracting layer properties from GDS and OASIS files through KLayout. * Calculating dies per wafer, edge exclusions, and yield projections across multiple wafer standard diameters via AnySilicon. * Automating full RTL-to-GDSII ASIC flows and standard cell placement through containerized OpenLane executions.

Key facts

• https://github.com/ssql2014/mcp4eda
• Developer Tools & Code Intelligence, Other & General Purpose
• eda, hardware-design, semiconductor, systemverilog, verilog

Part of MCP Servers

What is mcp4eda?

mcp4eda is a collection of Model Context Protocol servers designed for electronic design automation. It exposes tools like Yosys, Verilator, Verible, GTKWave, KLayout, AnySilicon, and OpenLane to LLM interfaces, allowing users to analyze, synthesize, simulate, and check hardware designs using plain language commands.

Which tools are included in mcp4eda?

The suite includes eight MCP servers covering synthesis via Yosys, simulation through Verilator, linting with Verible, waveform inspection via GTKWave, layout handling with KLayout, die yield calculations through AnySilicon, semiconductor supply chain analysis, and full RTL-to-GDSII runs with OpenLane.

Which MCP clients work with mcp4eda?

mcp4eda is designed to work with any client supporting the Model Context Protocol, including Claude Desktop on macOS, Windows, and Linux. The servers run locally over Node.js and communicate through standard input and output streams.

What prerequisites are needed to run mcp4eda?

You need Node.js version 18 or higher, npm, and the specific underlying EDA tools installed on your host system or inside Docker containers depending on which sub-servers you plan to configure and execute.