Open Patent MCP Gateway
Multi-source patent search MCP server that allows AI agents to search global patent databases (Google Patents, BigQuery, EPO OPS) via natural language queries.
README
Open Patent MCP Gateway
多源专利检索 MCP 网关 — 让 AI Agent 直接搜索全球专利数据库(Google Patents、BigQuery、EPO OPS)。
给你的 AI Agent 加上专利检索(30 秒)
在 GPT Work / Cursor / Claude Desktop 的 MCP 配置中添加:
{
"mcpServers": {
"open-patent": {
"url": "https://open-patent-mcp-gateway.the0shell0.workers.dev/mcp",
"headers": {
"Authorization": "Bearer dev-token-local"
}
}
}
}
然后直接对 AI 说:「帮我查一下 lithium battery 固态电解质相关的专利」即可。
可用的 MCP 工具
| 工具 | 用途 | 是否消耗配额 |
|---|---|---|
search_patents |
关键词/分类号/申请人/日期检索,支持 basic/standard/deep 三档 | 是 |
get_patent |
按公开号/申请号获取单条专利详情 | 否 |
get_family |
获取专利族成员和优先权链 | 否 |
get_citations |
获取引证文献(前引/后引,审查员/申请人) | 否 |
get_legal_status |
获取法律状态和法律事件 | 否 |
get_drawings |
获取附图元数据和图片 URL | 否 |
list_providers |
列出已接入的数据库及其能力 | 否 |
check_provider_status |
检查指定数据库的认证和可用性 | 否 |
get_quota_status |
查询各数据库的配额余量和重置时间 | 否 |
guide_setup |
引导用户独立部署(不接收 key,只返回步骤) | 否 |
检索档位
| 档位 | 数据源 | 配额消耗 | 适用场景 |
|---|---|---|---|
basic |
EPO OPS | 免费 | 快速查证、已知分类号检索 |
standard |
EPO OPS + Google Patents | 1 次 SerpApi | 探索性关键词搜索 |
deep |
全部 + BigQuery 全量历史 | 1 次 SerpApi + ~0.24 TiB | 穷尽式现有技术检索 |
想自己部署?(v0.2 双模式)
本项目支持两种使用模式:
- 共享模式(默认):直接用上方 30 秒配置接入维护者的 Worker 与 API,零配置。
- 独立模式(可选升级):注册自己的 API 与 Cloudflare,部署私有实例。BYOK,密钥只在你自己的 Worker 上。
需要独立部署时,任选一种引导方式:
| 方式 | 适合 | 入口 |
|---|---|---|
| 浏览器引导页 | 非技术用户 | 打开 https://open-patent-mcp-gateway.the0shell0.workers.dev/setup |
| AI 引导 | 在 GPT Work 等客户端里直接让 AI 帮你部署 | 对 AI 说「教我独立部署这个 MCP」→ 触发 guide_setup 工具 |
| CLI 向导 | 喜欢终端的用户 | npm run setup |
| 静态文档 | 离线参考 | docs/getting-api-keys.md |
四种形态共用同一份数据源 src/onboarding/steps.ts,内容始终一致。
独立模式最低只需 SerpApi 一个 key(standard 档位可用);EPO OPS 与 BigQuery 为可选增强。详细路线图见 docs/v0.2-roadmap.md。
git clone https://github.com/lgxycsg/open-patent-mcp-gateway.git
cd open-patent-mcp-gateway
npm install
npm run setup # 交互式向导:引导注册 + 验证 key + 生成 .dev.vars
npx wrangler deploy # 部署到你的 Cloudflare 账号
部署后得到你自己的 https://xxx.workers.dev,替换上方 30 秒配置中的 URL 即可。
项目结构
src/
├─ core/
│ ├─ mcp/server.ts # MCP server + 10 tools + HTTP/stdio transport
│ ├─ models/ # PatentRecord, SearchQuery, Identifier
│ ├─ errors/errors.ts # 15 error codes + secret redaction
│ ├─ routing/router.ts # Provider router (single/parallel/fallback)
│ ├─ normalization/number.ts # Publication number normalization
│ ├─ deduplication/index.ts # Dedup + family aggregation
│ └─ security/secrets.ts # Credential management + token validation
├─ onboarding/ # v0.2 独立模式引导(单一数据源 + 网页 + 验证)
│ ├─ steps.ts # 4 种引导形态共用的步骤数据源
│ ├─ setup-page.ts # Worker /setup 路由的 HTML 渲染器
│ └─ verify.ts # /setup/verify 服务端 key 验证(无状态)
├─ providers/
│ ├─ mock/index.ts # Mock provider (testing)
│ ├─ epo_ops/ # EPO OPS (official API, OAuth)
│ ├─ google_patents/ # Google Patents via SerpApi
│ ├─ bigquery/ # Google Patents Public Datasets
│ ├─ types.ts # Provider interface + Manifest
│ └─ registry.ts # Static provider registry
├─ worker.ts # Cloudflare Workers entry
└─ index.ts # Node.js entry
测试
npm run typecheck # 类型检查
npm test # 388 个单元测试
npm run build # 编译
许可证
MIT
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.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
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.