LTspice MCP

LTspice MCP for macOS

💡 User Note & SOP Context: This document serves as a personal Standard Operating Procedure (SOP) to streamline my own mental workflow and automation setup. No source code in this repository has been altered. Full credit for the engineering, logic, and core architecture belongs entirely to the original developers and the upstream projects credited below.

Model Context Protocol (MCP) server that enables agents and MCP clients to control LTspice on macOS for simulation, schematic generation, data extraction, verification, and rendering.

This project supports practical automation with reliable execution, reproducible artifacts, and outputs that closely match native LTspice behavior.

Credited to:

• gtnoble/ngspice-mcp

• luc-me/ltspiceMCP

• xuio/ltspice-mcp (core architecture and ScreenCaptureKit integration)

What You Can Do

• Run LTspice simulations via MCP (simulateNetlistFile, runSimulation, queue tools).

• Generate and refine schematics (createSchematic*, lint/clean/debug tools).

• Render authentic LTspice images for schematics, plots, and symbols.

• Query RAW vectors and derive analysis metrics (bandwidth, margins, rise/fall times, settling time).

• Automate .meas statements and assertion-driven verification workflows.

• Execute stepped and Monte Carlo studies with structured result outputs.

Related MCP server: Universal Netlist MCP Server

Quick Start

1. Prerequisites

• macOS with LTspice installed (default path: /Applications/LTspice.app).

• Python 3.11 or higher.

• uv (recommended) or pip.

2. Installation

uv sync

For pip:

python3 -m venv .venv
source .venv/bin/activate
pip install -e .

3. Start the MCP Daemon

./scripts/ltspice_mcp_daemon.sh start
./scripts/ltspice_mcp_daemon.sh status

Default endpoint: http://127.0.0.1:8765/mcp

4. Grant macOS Permissions (One-Time)

./scripts/ltspice_mcp_daemon.sh trigger-initial-permissions
./scripts/ltspice_mcp_daemon.sh check-accessibility

Claude Desktop Integration

The architecture uses Anthropic’s Model Context Protocol (MCP). Claude can actively call remote tools on your local machine via Server-Sent Events (SSE).

Configuration File

Edit the global configuration:

open -e ~/Library/Application\ Support/Claude/claude_desktop_config.json

Merge or replace the mcpServers section:

{
  "mcpServers": {
    "ltspice-mcp": {
      "command": "/opt/homebrew/bin/npx",
      "args": ["-y", "mcp-remote", "http://127.0.0.1:8765/mcp"]
    }
  }
}

Operating Instructions

Step 1: Start the Local Daemon
Before using Claude:

./scripts/ltspice_mcp_daemon.sh stop
./scripts/ltspice_mcp_daemon.sh start

Verify the process is listening on http://127.0.0.1:8765/mcp.

Step 2: Prompting Claude
Restart Claude Desktop (Cmd + Q). Confirm the plug/hammer icon appears. Use intent-driven prompts such as:

• AC Analysis:
  “Please use the local ltspice-mcp tool. Locate rc_lowpass_ac.asc in my common circuits directory. Run the AC sweep and extract a uniform decade-spaced subsample of the complex out-node vector data.”

• Transient Analysis:
  “Using component definitions from the prior AC analysis (tau ≈ 0.1 ms), generate a transient netlist for a step response from 0 to 1 ms. Run headlessly and return the interpolated output voltages.”

Step 3: Shut Down

./scripts/ltspice_mcp_daemon.sh stop

Client Configuration

URL-Based MCP Clients

[mcp_servers.ltspice]
url = "http://127.0.0.1:8765/mcp"
enabled = true

Claude Desktop (mcp-remote)

See the JSON configuration above.

Stdio (Subprocess) Mode

{
  "mcpServers": {
    "ltspice-mcp": {
      "command": "ltspice-mcp",
      "args": ["--transport", "stdio"],
      "cwd": "/absolute/path/to/ltspice-mcp"
    }
  }
}

Core Capabilities

Setup and Diagnostics

• getLtspiceStatus, getLtspiceUiStatus

• daemonDoctor, tailDaemonLog, getRecentErrors, getCaptureHealth

Simulation and Queueing

• simulateNetlist, simulateNetlistFile, runSimulation

• queueSimulationJob, listJobs, jobStatus, cancelJob, listJobHistory

Schematic Workflows

• createSchematic, createSchematicFromNetlist, createSchematicFromTemplate

• validateSchematic, lintSchematic, autoDebugSchematic

• inspectSchematicVisualQuality, autoCleanSchematicLayout

Data, Measurements, and Verification

• getPlotNames, getVectorsInfo, getVectorData, getLocalExtrema

• getBandwidth, getGainPhaseMargin, getRiseFallTime, getSettlingTime

• parseMeasResults, runMeasAutomation, runVerificationPlan, runSweepStudy

Native LTspice Rendering

• renderLtspiceSchematicImage

• renderLtspicePlotImage, renderLtspicePlotPresetImage

• renderLtspiceSymbolImage

• startLtspiceRenderSession, endLtspiceRenderSession

Daemon Operations

./scripts/ltspice_mcp_daemon.sh start
./scripts/ltspice_mcp_daemon.sh restart
./scripts/ltspice_mcp_daemon.sh stop
./scripts/ltspice_mcp_daemon.sh status
./scripts/ltspice_mcp_daemon.sh logs --lines 200
./scripts/ltspice_mcp_daemon.sh logs --follow

Permission helpers:

./scripts/ltspice_mcp_daemon.sh trigger-screen-recording-permission
./scripts/ltspice_mcp_daemon.sh trigger-accessibility-permission

Acknowledgments & Architecture Credits

This automation matrix adapts underlying client-server abstraction layers and local operating system hooks from public framework builds.

• Core Application Engine: Heavily extends and builds upon the native macOS tooling layout pioneered by xuio/ltspice-mcp. Special recognition to the original architecture for enabling ScreenCaptureKit hooks and automated multi-window script handlers.

• Protocol Core: Driven by the Anthropic Model Context Protocol (MCP) leveraging mcp-remote serialization bridges.