@car2db/mcp-server
v1.0.2
Published
MCP server for Car2DB API v3, allowing LLM agents to work with the automotive database
Downloads
226
Maintainers
Readme
Car2DB MCP Server
MCP (Model Context Protocol) server for Car2DB API v3, enabling LLM agents (Claude, ChatGPT, Cursor, VSCode) to directly interact with automotive database.
Features
- Complete Vehicle Database — Access detailed specifications for cars, motorcycles, and trucks
- Smart Search — Natural language vehicle search
- Rich Data — Technical specs, equipment packages, and detailed characteristics
- Multi-language — Support for 11 languages
- Dual Transport — stdio (Claude Desktop/Cursor) and SSE (remote access)
- Docker Ready — Production-ready containerization
Quick Start
Claude Desktop Configuration
Add to your claude_desktop_config.json:
{
"mcpServers": {
"car2db": {
"command": "npx",
"args": ["-y", "@car2db/mcp-server"],
"env": {
"CAR2DB_API_KEY": "your_api_key_here",
"CAR2DB_REFERER": "https://yourproject.com",
"CAR2DB_LANGUAGE": "en-US"
}
}
}
}Get your API key: https://car2db.com/api/ Try demo key: https://car2db.com/api-token-demo/
GitHub Copilot Configuration
For GitHub Copilot, use the .vscode/mcp.json configuration file in your project:
{
"servers": {
"car2db": {
"command": "node",
"args": ["dist/index.js"],
"env": {
"CAR2DB_API_KEY": "your_api_key_here",
"CAR2DB_REFERER": "https://yourproject.com",
"CAR2DB_LANGUAGE": "en-US",
"MCP_TRANSPORT": "stdio"
}
}
}
}Or use npx for zero-install:
{
"mcpServers": {
"car2db": {
"command": "npx",
"args": ["-y", "@car2db/mcp-server"],
"env": {
"CAR2DB_API_KEY": "your_api_key_here",
"CAR2DB_REFERER": "https://yourproject.com",
"CAR2DB_LANGUAGE": "en-US"
}
}
}
}Get your API key: https://car2db.com/api/
Try demo key: https://car2db.com/api-token-demo/
Installation Options
1. MCPB Package (One-Click Install)
Download the pre-built MCPB extension for easy installation in Claude Desktop or other MCP clients:
Download car2db-api-v3-mcp.mcpb
MCPB is a ZIP archive containing a local MCP server and manifest.json describing the server and its capabilities — just download and install with one click.
Installation Steps:
- Download the
.mcpbfile from GitHub releases - Double-click the file or drag it into your MCP client
- Configure your API key in the setup dialog (optional, demo key will be used if not provided)
- Click "Install" and start using the extension
GitHub Releases: https://github.com/car2db/car2db-api-v3-mcp/releases
2. npx (Zero Install)
npx @car2db/mcp-server3. Global npm Installation
npm install -g @car2db/mcp-server
car2db-mcp4. Local Development
git clone https://github.com/car2db/car2db-api-v3-mcp.git
cd car2db-api-v3-mcp
npm install
npm run build
npm start5. Docker (for SSE mode)
docker-compose up -dOr manually:
docker build -t car2db-mcp .
docker run -p 3000:3000 \
-e CAR2DB_API_KEY=your_key \
-e CAR2DB_REFERER=https://yourproject.com \
car2db-mcpConfiguration
Environment Variables
| Variable | Required | Default | Description |
|----------|----------|---------|-------------|
| CAR2DB_API_KEY | ✅ | — | Bearer token for API authorization. Get it at car2db.com/api |
| CAR2DB_REFERER | ✅ | — | URL of your project/site (e.g. https://yourproject.com). Must match the URL registered with your API key. |
| CAR2DB_LANGUAGE | ❌ | en-US | API response language: en-US, ru-RU, de-DE, es-ES, it-IT, pl-PL, fr-FR, da-DA, lv-LV, th-TH, zh-CN |
| MCP_TRANSPORT | ❌ | stdio | Transport mode: stdio (Claude Desktop/Cursor) or sse (HTTP Server-Sent Events) |
| MCP_SSE_PORT | ❌ | 3000 | Port for SSE mode (only used when MCP_TRANSPORT=sse) |
Referer header
The Car2DB API requires a Referer header in every request. Set CAR2DB_REFERER to the URL of the project or site that uses this MCP server. This URL must match the one registered with your API key on car2db.com.
If CAR2DB_REFERER is not set, a warning is printed at startup and API requests will fail with 403 Forbidden.
Example .env file
CAR2DB_API_KEY=your_api_key_here
CAR2DB_REFERER=https://yourproject.com
CAR2DB_LANGUAGE=en-US
MCP_TRANSPORT=stdioAvailable Tools
Smart Tools (High-Level)
| Tool | Description | Key Parameters |
|------|-------------|----------------|
| search_vehicles | Natural language vehicle search | q (search text), typeId, yearFrom, yearTo |
| get_trim_full | Complete trim specifications | trimId |
| get_equipment_full | Full equipment/package details | equipmentId |
| browse_catalog | Navigate catalog hierarchy | level (makes/models/generations/series/trims/equipments), parent IDs |
Low-Level Tools (Direct API Access)
| Tool | Description | Key Parameters |
|------|-------------|----------------|
| list_makes | List all vehicle manufacturers | page, itemsPerPage, typeId |
| list_models | List models by make | makeId, typeId, pagination |
| list_generations | List model generations | modelId, typeId, pagination |
| list_series | List series (body styles) | modelId, generationId, typeId, pagination |
| list_trims | List trims (engine variants) | seriesId, modelId, typeId, pagination |
| get_year_vehicles | Vehicles by production year | year, typeId |
Tool Parameters Reference
typeId: Vehicle type filter1= Cars2= Motorcycles3= Trucks
page: Page number (default: 1)itemsPerPage: Items per page (default: 30)
Usage Examples
Example 1: Find specific vehicle specifications
Claude: "What are the specifications of 2020 BMW X5 xDrive40i?"
Agent workflow:
1. search_vehicles(q: "BMW X5 xDrive40i 2020")
2. get_trim_full(trimId: <from_search_results>)
3. Returns: breadcrumbs, key specs, full specifications, available equipmentsExample 2: Compare equipment packages
Claude: "Compare equipment packages for Toyota Camry 2023"
Agent workflow:
1. search_vehicles(q: "Toyota Camry 2023")
2. get_trim_full(trimId: ...) → get equipment list
3. get_equipment_full(equipmentId: A) → package A options
4. get_equipment_full(equipmentId: B) → package B options
5. Compare and present differencesExample 3: Browse catalog
Claude: "Show me all Tesla models"
Agent workflow:
1. browse_catalog(level: "makes") → find Tesla
2. browse_catalog(level: "models", makeId: 56) → all Tesla models
3. browse_catalog(level: "trims", modelId: 789) → trims for Model 3Example 4: Year-based search
Claude: "What electric SUVs were available in 2024?"
Agent workflow:
1. get_year_vehicles(year: 2024, typeId: 1)
2. Filter for electric powertrains
3. Filter for SUV body types
4. Present resultsSSE Mode (Remote Access)
For remote access or integration with web services, use SSE transport:
Start SSE Server
export MCP_TRANSPORT=sse
export MCP_SSE_PORT=3000
npm startOr with Docker:
docker-compose up -dSSE Endpoints
- SSE Connection:
GET http://localhost:3000/sse - Messages:
POST http://localhost:3000/messages - Health Check:
GET http://localhost:3000/health
API Coverage
This MCP server provides access to Car2DB API v3 endpoints:
/search/vehicles— Vehicle search/makes— Manufacturers list/models— Vehicle models/generations— Model generations/series— Body style series/trims— Engine/transmission trims/trims/{id}/full— Complete trim data/equipments/{id}/full— Equipment packages/years/{year}— Vehicles by year
Base URL: https://v3.api.car2db.com
Development
Build
npm run buildType Check
npm run typecheckDevelopment Mode
npm run devGet API Access
- Purchase API Key: https://car2db.com/api/
- Try Demo Key: https://car2db.com/api-token-demo/
- API Documentation: https://v3.api.car2db.com/docs
Requirements
- Node.js >= 20.0.0
- Valid Car2DB API key
License
MIT License
Links
- Repository: https://github.com/car2db/car2db-api-v3-mcp
- npm Package: https://www.npmjs.com/package/@car2db/mcp-server
- Car2DB API: https://car2db.com/api/
- Model Context Protocol: https://modelcontextprotocol.io/
Support
For API-related questions and support:
- Email: [email protected]
- Website: https://car2db.com
- Telegram: https://t.me/car2db_support_bot
For MCP server issues:
- GitHub Issues: https://github.com/car2db/car2db-api-v3-mcp/issues
