partsgeek-mcp-server
Enables searching PartsGeek for parts, retrieving product details, and obtaining cart estimates for price comparison, without placing orders or requiring a PartsGeek account.
README
partsgeek-mcp-server
Notice: Open-source local tooling provided AS IS by the authors and contributors; not affiliated with ALLDATA, AutoZone, PartsGeek, or any OEM—use only with accounts you control and at your own risk (see DISCLAIMER.md).
License: MIT
Local MCP server for public catalog / cart estimate workflows on PartsGeek. Designed for price comparison against shop commercial pricing—not for placing orders.
Safety hard rules
- Never checkout. No PayPal, Fastlane, cards, or Place Order.
- Cart page only (
/cart/) for shipping / fee estimates. - Do not open
/checkout/or payment UI.
What it does
| Tool | Purpose |
|---|---|
search_parts |
Y/M/M + keyword → hits with clickable URLs + prices |
get_product |
Product detail (SKU, fitment, ships note, url) |
cart_estimate |
Add product URLs to cart; return subtotal + expedited + protection + S&H |
clear_browser |
Drop Playwright session / cookies |
Default ship-to note for estimates (override when calling the tool):
18440 Paulson Drive, Port Charlotte, FL 33954 (example shop address—change for your shop).
Requirements
- Python ≥ 3.10
- Playwright Chromium
- No PartsGeek account required for basic search/cart estimate
Install
git clone <this-repo-url> partsgeek-mcp-server
cd partsgeek-mcp-server
pip install -e .
playwright install chromium
# ~/.grok/config.toml
[mcp_servers.partsgeek]
command = "partsgeek-mcp"
enabled = true
[mcp_servers.partsgeek.env]
PARTSGEEK_HEADLESS = "1"
PARTSGEEK_TIMEOUT = "180"
hermes mcp add partsgeek --command partsgeek-mcp \
--env PARTSGEEK_HEADLESS=1 PARTSGEEK_TIMEOUT=180
Authentication
PartsGeek catalog browsing does not require a shop login. The tool drives a local headless browser on your machine for search/cart DOM only.
There is no import_browser_cookies step for PartsGeek accounts. If you ever log into PartsGeek in that browser profile, treat it like any other local session and do not commit cookie jars.
Data handling
- Runs entirely on your device (Playwright + HTTPS to PartsGeek).
- Authors do not operate a credential-collection backend.
- Full terms: DISCLAIMER.md.
Pair with shop quoting
Use azpro for commercial cost/list and alldata for labor. The shop-quote skill documents dual-source RO format and requires clickable product links for PartsGeek lines.
Development
pytest -q
License
MIT — see LICENSE.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。