Purpose
BoltPatternHQ MCP Server is the official Model Context Protocol (MCP) server provided by BoltPatternHQ, used to connect standardized, OEM-verified automotive wheel engineering specification data within AI coding assistants (Claude, Cursor, Windsurf, VS Code, etc.), preventing large language models from fabricating (hallucinating) vehicle wheel parameters when developing automotive-related applications.
Main Capabilities
The server exposes 3 standardized AI tools via stdio:
- search_bolt_pattern: Search wheel specifications by manufacturer brand, model name, chassis code, year, or exact PCD string (such as "5x114.3"), returning a detailed fitment report including center bore, offset tolerance, and OEM wheel notes. Input examples: {"query": "Jimny"}, {"query": "5x114.3"}, {"query": "Roadster"}.
- get_make_models: List all supported automotive brand manufacturers and key models in the public MCP library, outputting a Markdown overview organized by category.
- get_sample_data: Export the complete curated JSON sample dataset, which can be directly imported into CAD workflows, SQL databases, or unit tests.
Standardized engineering metrics include: PCD (such as 4x100, 5x114.3, 5x139.7), center bore (mm), offset range (ET, mm), lug nut thread specifications (M12x1.5, M14x1.25, ball/cone seat notes), and bolt count. The built-in curated sample library covers classic JDM, European, and American models (Mazda MX-5 Miata, Suzuki Jimny, Toyota Raize/Rocky, BMW 3 Series, Volkswagen Golf GTI, Ford Mustang, Tesla Model 3/Y, etc.), supporting 0ms offline queries.
Usage
It can be run directly via Node.js and npx, with no need to install a Python virtual environment.
Claude Desktop configuration: Add the following configuration to the Claude Desktop configuration file: macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json
Cursor configuration: Open Settings -> Features -> MCP Servers -> Add new MCP server: Name: boltpatternhq Command: npx -y boltpatternhq-mcp
Smithery CLI: Can be run or installed via Smithery.
VS Code (Cline / Roo Code / Continue) and Windsurf: Add to the workspace or global MCP settings JSON.
It can also be cloned from source and built and run locally. The project is open source under the MIT license (server wrapper and public sample data).
Use Cases
- Developing automotive e-commerce stores (Shopify / WooCommerce)
- Wheel fitment calculators
- CAD part models
- ERP systems
- AI coding debugging and prototyping that require wheel specification data
Limitations
This MCP server only provides instant access to the BoltPatternHQ public sample library, for prototyping and AI debugging. Production-grade e-commerce websites, global tire/wheel export businesses, and commercial parts lookup systems require the full master database.
Commercial License and Master Dataset
The full master database includes: 11,601 verified model-year records (1992–2026), 66 global automotive manufacturers (covering JDM, USDM, European, Australian, and electric platforms), provided in standardized UTF-8 CSV and JSON formats, ready for immediate import into SQL/ERP. It uses a one-time perpetual license, priced at $29, with no monthly subscription API fees.
Other Information
The project has been adapted and verified for inclusion in multiple AI ecosystem registries: Official MCP Registry (.mcp/server.json), Glama.ai MCP Servers, Awesome MCP Servers. Official website and live search engine:https://boltpatternhq.com/ MCP and API documentation:https://boltpatternhq.com/data/ Full dataset access:https://boltpatternhq.com/data/ Developer blog:https://double2.hatenablog.com/
