DCS Lua Runner MCP Server
Enables AI assistants to interact with DCS World in real-time through chat by executing Lua scripts and providing tools for mission info, unit management, coordinate conversion, and more.
README
DCS Lua Runner MCP Server
An MCP (Model Context Protocol) server that enables AI assistants to interact with DCS World (Digital Combat Simulator) in real-time through chat.
Features
This MCP server provides the following tools for AI interaction with DCS:
Tools Available
-
execute_lua - Execute arbitrary Lua code on DCS server
- Execute custom Lua scripts
- Choice of mission or GUI environment
- Returns execution results
-
get_mission_info - Get current mission information
- Mission time
- Theatre/map name
- Mission date
-
get_player_info - Get player aircraft information
- Aircraft name and type
- Position (x, y, z coordinates)
- Altitude
- Speed
- Heading
-
get_all_units - List all units in the mission
- Filter by coalition (red, blue, neutral, or all)
- Unit names, types, positions
- Coalition affiliations
-
spawn_unit - Spawn new units in the mission
- Ground units or vehicles
- Specify position, heading, coalition
- Custom unit names
-
send_message - Display messages in DCS
- Send text messages to all players
- Configurable display duration
-
get_theatre_info - Get theatre/map information
- Current map name
- Theatre details
-
get_aircraft_list - List all aircraft in mission
- All airborne units
- Positions and altitudes
- Coalition information
-
convert_coordinates - Convert real-world coordinates to DCS
- Convert Lat/Long (decimal degrees) to DCS X/Z coordinates
- Optional altitude parameter
- Uses the theatre's native
coord.LLtoLODCS API
-
convert_dcs_to_ll - Convert DCS coordinates to Lat/Long
- Convert DCS X/Z to real-world Latitude/Longitude
- Optional Y/altitude parameter
- Uses the theatre's native
coord.LOtoLLDCS API
Prerequisites
- DCS World installed with DCS Fiddle server running
- Node.js installed (for running the MCP server)
Configuration
The MCP server reads settings from dcs_lua_runner_settings.json in the following priority order:
- Environment Variable (if set): Path specified in
DCS_SETTINGS_PATH - MCP Server Directory (default):
dcs_lua_runner_settings.jsonin the same folder as the MCP server - Fallback Defaults: If no settings file is found, uses default localhost configuration
Quick Setup
-
Copy the template file:
cp dcs_lua_runner_settings.json.template dcs_lua_runner_settings.json -
Edit the settings file with your DCS server details (the defaults work for most local setups)
The MCP server will automatically find and load the settings file from its own directory.
Settings Used
server_address- DCS server address for remote connectionsserver_port- Port for mission environment (default: 12080)server_address_gui- GUI environment server addressserver_port_gui- Port for GUI environment (default: 12081)use_https- Whether to use HTTPSweb_auth_username- Username for authenticationweb_auth_password- Password for authenticationrun_code_locally- Execute on local server (127.0.0.1)run_in_mission_env- Execute in mission vs GUI environment
Installation
Method 1: Direct Installation (Pre-built)
The release zip already contains the pre-built build/index.js — no Node.js build step required.
-
Extract the release zip (e.g.
dcs-lua-runner-mcp-v1.1.0.zip) to your preferred location, e.g.:C:\dcs-lua-runner-mcp\ -
Run the installer script — open PowerShell, navigate to the extracted folder and run:
.\install-mcp-agent.ps1The script walks you through two steps automatically:
Step 1 — DCS Server Settings
- Asks whether DCS runs on the same machine (local) or on a remote IP
- Prompts for port, username, and password (current/template values shown in brackets — press Enter to keep)
- Saves the result to
dcs_lua_runner_settings.json - Also patches
dcs-fiddle-server.luawith IP, port, username and password — no need to edit it separately
Step 2 — Agent Registration
- Shows a numbered menu of supported agents with their config status
- Enter one or more numbers separated by commas or spaces, or type
all - Type
skip(or press Enter with no input) to skip agent registration entirely - Patches each selected agent config file (creates the file if it doesn't exist yet)
1. Claude Desktop 2. Cline (VS Code extension) 3. GitHub Copilot Chat (VS Code) 4. Cursor 5. Windsurf 6. Claude Code CLI 7. GitHub Copilot CLITip: You can skip the interactive prompts entirely with flags:
# Configure specific agents (e.g. Copilot Chat + Cline) .\install-mcp-agent.ps1 -AgentIds 2,3 # Configure all agents at once .\install-mcp-agent.ps1 -All # Skip agent registration (only update settings + fiddle script) .\install-mcp-agent.ps1 -SkipAgents # Point to a build in a custom location .\install-mcp-agent.ps1 -ServerPath "D:\tools\dcs-mcp\build\index.js" -
Install the DCS server script — copy
dcs-fiddle-server.luainto your DCS mission scripts or into the DCSScripts\Hooksfolder and load it (see DCS Server Script Installation). -
Restart your agent (reload VS Code, restart Claude Desktop, etc.) to apply the configuration changes.
Method 2: Installation from GitHub
-
Clone the repository:
git clone https://github.com/sevenfifty777/dcs-lua-runner-mcp.git cd dcs-lua-runner-mcp -
Install dependencies:
npm install -
Build the project:
npm run build -
Configure settings and register agents:
Option A — Automatic (recommended): Run the installer script, which handles everything in one go:
.\install-mcp-agent.ps1The script will:
- Prompt for DCS server settings and save
dcs_lua_runner_settings.json - Patch
dcs-fiddle-server.luawith the same values automatically - Let you pick which agents to register from a numbered menu (type
skipto skip this step)
Option B — Manual:
Copy the template and fill in your DCS server details:
cp dcs_lua_runner_settings.json.template dcs_lua_runner_settings.jsonEdit
dcs_lua_runner_settings.json:{ "server_address": "127.0.0.1", "server_port": 12080, "server_address_gui": "127.0.0.1", "server_port_gui": 12081, "use_https": false, "web_auth_username": "your_username", "web_auth_password": "your_password", "run_code_locally": true, "run_in_mission_env": true, "return_display_format": "lua" }Then update your agent's MCP config file to point to
build/index.js. Placedcs_lua_runner_settings.jsonin the same directory as the MCP server so it is found automatically.For Cline (VS Code) — edit
%APPDATA%\Code\User\globalStorage\saoudrizwan.claude-dev\settings\cline_mcp_settings.json:{ "mcpServers": { "dcs-lua-runner-mcp": { "command": "node", "args": ["C:/absolute/path/to/dcs-lua-runner-mcp/build/index.js"] } } }For Claude Desktop — edit
%APPDATA%\Claude\claude_desktop_config.json(Windows) or~/Library/Application Support/Claude/claude_desktop_config.json(macOS):{ "mcpServers": { "dcs-lua-runner-mcp": { "command": "node", "args": ["C:/absolute/path/to/dcs-lua-runner-mcp/build/index.js"] } } }For GitHub Copilot Chat (VS Code) — edit
%APPDATA%\Code\User\mcp.json:{ "servers": { "dcs-lua-runner-mcp": { "command": "node", "args": ["C:/absolute/path/to/dcs-lua-runner-mcp/build/index.js"], "type": "stdio" } } }For GitHub Copilot CLI — edit
%USERPROFILE%\.copilot\mcp-config.json:{ "mcpServers": { "dcs-lua-runner-mcp": { "command": "node", "args": ["C:/absolute/path/to/dcs-lua-runner-mcp/build/index.js"], "type": "stdio" } } }Custom settings path — if
dcs_lua_runner_settings.jsonlives elsewhere, pass its path via environment variable:{ "mcpServers": { "dcs-lua-runner-mcp": { "command": "node", "args": ["C:/absolute/path/to/dcs-lua-runner-mcp/build/index.js"], "env": { "DCS_SETTINGS_PATH": "C:/custom/path/to/dcs_lua_runner_settings.json" } } } } - Prompt for DCS server settings and save
-
Install the DCS server script — copy
dcs-fiddle-server.luainto your DCS mission scripts or into the DCSScriptsfolder and load it (see DCS Server Script Installation). -
Restart your agent (reload VS Code, restart Claude Desktop, etc.)
-
Test the connection:
Ask your AI assistant: "What's the current mission time in DCS?"
Verification
After installation, verify the MCP server is working:
- Check that
build/index.jsexists - Ensure DCS World is running with DCS Fiddle server
- Ask your AI assistant to get mission info
- Check MCP server logs in your client for any errors
Usage Examples
Once the MCP server is running, you can ask your AI assistant to interact with DCS:
Get Mission Information
"What's the current mission time in DCS?"
"What theatre/map am I flying in?"
Get Player Information
"Where is my aircraft?"
"What's my current altitude and speed?"
Spawn Units
"Spawn a T-72B tank at coordinates x=100000, z=200000 for the red coalition"
"Create an M-1 Abrams at position 150000, 250000 facing north"
Send Messages
"Send a message to all players saying 'Mission starting in 5 minutes'"
Execute Custom Lua
"Execute this Lua code in DCS: return timer.getTime()"
"Run this script to get all blue coalition units"
List Units and Aircraft
"Show me all units in the mission"
"List all red coalition aircraft"
"What aircraft are currently in the mission?"
Convert Coordinates
"Convert latitude 41.123, longitude 44.987 to DCS coordinates"
"What are the DCS X/Z coordinates for lat 51.5, lon 37.2?"
"Convert DCS coordinates X=100000, Z=200000 to latitude/longitude"
"What is the real-world position of the unit at DCS X=50000, Z=150000?"
DCS Setup Requirements
⚠️ Security Notice: Setting up DCS Fiddle requires disabling DCS sandboxing and, optionally, exposing a network port. Read DCS_Setup_Risks.md for a full risk assessment and mandatory security measures before proceeding.
1. DCS Fiddle Server Installation
Step 1 — Locate your Saved Games folder
DCS uses a folder under Windows Saved Games to store user scripts. The exact name depends on your DCS branch:
| DCS Branch | Saved Games Folder |
|---|---|
| DCS World (stable) | %USERPROFILE%\Saved Games\DCS |
| DCS World OpenBeta | %USERPROFILE%\Saved Games\DCS.openbeta |
Open File Explorer and navigate to the correct path, e.g.:
C:\Users\<YourUser>\Saved Games\DCS
Step 2 — Create the Hooks folder (if it does not exist)
Inside your Saved Games DCS folder, create the following folder structure if it is not already there:
Saved Games\DCS\Scripts\Hooks\
On Windows you can do this from PowerShell:
New-Item -ItemType Directory -Force -Path "$env:USERPROFILE\Saved Games\DCS\Scripts\Hooks"
For OpenBeta replace
DCSwithDCS.openbeta.
Step 3 — Copy the server script
Copy dcs-fiddle-server.lua from this repository into the Hooks folder:
Copy-Item ".\dcs-fiddle-server.lua" "$env:USERPROFILE\Saved Games\DCS\Scripts\Hooks\"
The final path should look like:
%USERPROFILE%\Saved Games\DCS\Scripts\Hooks\dcs-fiddle-server.lua
Step 4 — Verify the script is loaded
After launching DCS, open %USERPROFILE%\Saved Games\DCS\Logs\dcs.log and search for:
DCS Fiddle successfully initialized
If you see this line the server is running and listening on ports 12080 (mission environment) and 12081 (GUI environment).
Note: The Hooks folder is loaded by DCS at startup for both the main menu (GUI environment) and in-mission. You do not need to add anything to
autoexec.cfg.
2. DCS Desanitization
⚠️ Security Warning: This step disables DCS sandboxing and grants scripts full system access. See DCS_Setup_Risks.md for the full risk details and how to restore security afterwards.
Edit DCS_INSTALL\Scripts\MissionScripting.lua and comment out these two lines:
-- _G['require'] = nil -- comment this out
_G['loadlib'] = nil
-- _G['package'] = nil -- comment this out
3. Start DCS
- Launch DCS World
- The DCS Fiddle server will start automatically
- Verify the server is running (check DCS.log for "DCS Fiddle successfully initialized")
Troubleshooting
Connection Errors
Error: "Connection refused - check if DCS Fiddle server is running"
- Solution: Ensure DCS is running and DCS Fiddle server is installed correctly
Error: "Request timeout - check server connection"
- Solution: Check if DCS Fiddle server is bound to the correct address/port
- Verify firewall settings if using remote connections
Authentication Errors
Error: Unauthorized (401)
- Solution: Check username/password in
dcs_lua_runner_settings.json - Ensure authentication settings match DCS Fiddle server configuration
Execution Errors
Error: "No player found"
- Solution: Ensure you're in a mission with a player-controlled aircraft
- Try switching between mission and GUI environments
Error: Lua execution errors
- Solution: Check Lua syntax and DCS API availability
- Some functions only work in mission environment
Testing with MCP Inspector
The MCP Inspector is a developer tool that allows you to test and debug your MCP server interactively before integrating it with an AI client.
Running the Inspector
-
Ensure the project is built:
npm run build -
Start the MCP Inspector:
npm run inspector -
Using the Inspector:
- The inspector will open in your default web browser
- You'll see an interactive GUI showing:
- Available Tools: List of all 10 DCS interaction tools
- Tool Parameters: Input fields for each tool's parameters
- Request/Response: Real-time display of MCP communication
- Test Results: Output from each tool execution
-
Testing Individual Tools:
- Select a tool from the list (e.g.,
get_mission_info) - Fill in required parameters (if any)
- Click "Execute" to test the tool
- View the response in JSON format
- Select a tool from the list (e.g.,
-
Example Test Scenarios:
- Test connection: Use
get_mission_infoto verify DCS connection - Test Lua execution: Use
execute_luawith simple code likereturn "Hello from DCS" - Test spawning: Use
spawn_unitwith test coordinates - Debug errors: View detailed error messages and stack traces
- Test connection: Use
Inspector Benefits
- No AI client needed: Test without configuring Cline or Claude Desktop
- Interactive debugging: See requests and responses in real-time
- Parameter validation: Verify tool inputs before integration
- Quick iteration: Test changes immediately after rebuilding
Notes
- The inspector runs the MCP server in a test environment
- You still need DCS World running with the Fiddle server for actual DCS interaction
- The inspector uses the same
dcs_lua_runner_settings.jsonconfiguration
Development
Building from Source
cd C:\{your_path}\dcs-lua-runner-mcp
npm install
npm run build
Project Structure
dcs-lua-runner-mcp/
├── src/
│ └── index.ts # Main MCP server implementation
├── build/
│ └── index.js # Compiled JavaScript
├── package.json
├── tsconfig.json
└── README.md
Adding New Tools
To add new DCS interaction tools:
- Add tool definition to
ListToolsRequestSchemahandler - Implement tool logic in
CallToolRequestSchemahandler - Rebuild:
npm run build
Security Considerations
- Local Execution: By default, executes on
127.0.0.1(localhost only) - Remote Execution: Requires authentication (username/password)
- Code Validation: All Lua code is base64-encoded before transmission
- DCS Security: Requires DCS desanitization (understand the risks!)
License
MIT License - Based on the DCS Lua Runner GUI project
Credits
- Original DCS Fiddle: JonathanTurnock and john681611
- DCS Lua Runner VSCode Extension: omltcat
- GUI Implementation: Created for standalone Windows application
- Model Context Protocol: MCP SDK by Anthropic
Support
For issues related to:
- MCP Server: Check the build output and MCP settings
- DCS Connection: Verify DCS Fiddle server installation
- Lua Execution: Check DCS.log for detailed error messages
LobeHub
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.