Google Patents MCP Server
Google Patents MCP-Server ( google-patents-mcp )
Dieses Projekt bietet einen Model Context Protocol (MCP)-Server, der die Suche nach Google Patents-Informationen über die SerpApi Google Patents API ermöglicht.
Installation über Smithery
So installieren Sie den Google Patents MCP Server für Claude Desktop automatisch über Smithery :
npx -y @smithery/cli install @KunihiroS/google-patents-mcp --client claudeÄnderungsprotokoll
v0.2.0 (17.04.2025)
Fix: Leere Handler für die MCP-Methoden
resources/listundprompts/listimplementiert.Fix: Deklarierte
promptsbei der Serverinitialisierung.Aufgabe: Abhängigkeiten aktualisiert.
Diese Änderungen zielen darauf ab, die Kompatibilität mit MCP-Clients wie Claude Desktop zu verbessern, die diese Standardendpunkte möglicherweise benötigen, obwohl noch keine direkten Tests mit Claude Desktop durchgeführt wurden.
Related MCP server: Serper MCP Server
Merkmale
Bietet ein MCP-Tool
search_patentszum Durchsuchen von Google Patents.Verwendet SerpApi als Backend.
Kann ohne lokale Installation direkt mit
npxausgeführt werden.
Voraussetzungen
Node.js: Version 18 oder höher wird empfohlen.
npm: Erforderlich, um den
npx-Befehl auszuführen.SerpApi-API-Schlüssel: Sie benötigen einen gültigen API-Schlüssel von SerpApi , um die Google Patents API zu verwenden.
Schnellstart (mit npx)
Am einfachsten lässt sich dieser Server mit npx ausführen. Dieser Befehl lädt den Server (falls erforderlich) herunter und führt ihn direkt aus.
npx @kunihiros/google-patents-mcpHinweis: Ersetzen Sie @kunihiros/google-patents-mcp durch den tatsächlich veröffentlichten Paketnamen, falls dieser abweicht.
Der Server wird gestartet und wartet auf MCP-Anfragen auf der Standardeingabe/-ausgabe.
Konfiguration
Der Server benötigt Ihren SerpApi-API-Schlüssel. Sie können ihn auf eine der folgenden Arten bereitstellen:
Umgebungsvariable (empfohlen für MCP-Hosts): Setzen Sie die Umgebungsvariable
SERPAPI_API_KEYbeim Ausführen des Servers. MCP-Hostkonfigurationen ermöglichen häufig das Setzen von Umgebungsvariablen für Server.Beispiel für einen MCP-Host-Konfigurationsausschnitt (
config.jsonoder ähnlich):{ "mcpServers": { "google-patents-mcp": { "command": "npx", "args": [ "-y", // Skips confirmation if the package isn't installed locally "@kunihiros/google-patents-mcp" // Use the correct package name ], "env": { "SERPAPI_API_KEY": "YOUR_ACTUAL_SERPAPI_KEY" // Optional: Set log level // "LOG_LEVEL": "debug" } } } }.env-Datei: Erstellen Sie eine
.envDatei in dem Verzeichnis, in dem Sie dennpxBefehl ausführen (für lokale Tests oder wenn Sie keinen MCP-Host verwenden), oder in Ihrem Home-Verzeichnis (~/.google-patents-mcp.env), mit dem folgenden Inhalt:SERPAPI_API_KEY=YOUR_ACTUAL_SERPAPI_KEY # Optional: Set log level (e.g., debug, info, warn, error) # LOG_LEVEL=debugHinweis: Die Verwendung einer
.envDatei ist zwar für lokale Tests praktisch, für die Produktion oder die Integration mit MCP-Hosts ist jedoch das direkte Setzen der Umgebungsvariablen über die Hostkonfiguration der empfohlene und sicherere Ansatz. Der primäre Anwendungsfall ist die Ausführung übernpx, wobei Umgebungsvariablen typischerweise vom aufrufenden Prozess oder MCP-Host verwaltet werden.
Der Server sucht in der folgenden Reihenfolge nach .env Dateien: * ./.env (relativ zum Ausführungsort von npx ) * ~/.google-patents-mcp.env (in Ihrem Home-Verzeichnis)
Mitgeliefertes MCP-Tool
search_patents
Durchsucht Google Patents über SerpApi.
Eingabeschema:
{
"type": "object",
"properties": {
"q": {
"type": "string",
"description": "Search query (required). Although optional in SerpApi docs, a non-empty query is practically needed. Use semicolon (;) to separate multiple terms. Advanced syntax like '(Coffee) OR (Tea);(A47J)' is supported. See 'About Google Patents' for details."
},
"page": {
"type": "integer",
"description": "Page number for pagination (default: 1).",
"default": 1
},
"num": {
"type": "integer",
"description": "Number of results per page (default: 10). **IMPORTANT: Must be 10 or greater (up to 100).**",
"default": 10,
"minimum": 10,
"maximum": 100
},
"sort": {
"type": "string",
"enum": ["relevance", "new", "old"],
"description": "Sorting method. 'relevance' (default), 'new' (newest by filing/publication date), 'old' (oldest by filing/publication date).",
"default": "relevance"
},
"before": {
"type": "string",
"description": "Maximum date filter (e.g., 'publication:20231231', 'filing:20220101'). Format: type:YYYYMMDD where type is 'priority', 'filing', or 'publication'."
},
"after": {
"type": "string",
"description": "Minimum date filter (e.g., 'publication:20230101', 'filing:20220601'). Format: type:YYYYMMDD where type is 'priority', 'filing', or 'publication'."
},
"inventor": {
"type": "string",
"description": "Filter by inventor names. Separate multiple names with a comma (,)."
},
"assignee": {
"type": "string",
"description": "Filter by assignee names. Separate multiple names with a comma (,)."
},
"country": {
"type": "string",
"description": "Filter by country codes (e.g., 'US', 'WO,JP'). Separate multiple codes with a comma (,)."
},
"language": {
"type": "string",
"description": "Filter by language (e.g., 'ENGLISH', 'JAPANESE,GERMAN'). Separate multiple languages with a comma (,). Supported: ENGLISH, GERMAN, CHINESE, FRENCH, SPANISH, ARABIC, JAPANESE, KOREAN, PORTUGUESE, RUSSIAN, ITALIAN, DUTCH, SWEDISH, FINNISH, NORWEGIAN, DANISH."
},
"status": {
"type": "string",
"enum": ["GRANT", "APPLICATION"],
"description": "Filter by patent status: 'GRANT' or 'APPLICATION'."
},
"type": {
"type": "string",
"enum": ["PATENT", "DESIGN"],
"description": "Filter by patent type: 'PATENT' or 'DESIGN'."
},
"scholar": {
"type": "boolean",
"description": "Include Google Scholar results (default: false).",
"default": false
}
},
"required": ["q"]
}Ausgabe:
Gibt ein JSON-Objekt mit den Suchergebnissen von SerpApi zurück. Die Struktur folgt dem SerpApi-Antwortformat.
Beispielverwendung (MCP-Anforderung):
{
"mcp_version": "1.0",
"type": "CallToolRequest",
"id": "req-123",
"server_name": "google-patents-mcp",
"params": {
"name": "search_patents",
"arguments": {
"q": "organic light emitting diode",
"num": 10,
"language": "ENGLISH",
"status": "GRANT",
"after": "publication:20230101"
}
}
}Entwicklung
Klonen Sie das Repository (falls für die Entwicklung erforderlich):
# git clone <repository-url> # cd google-patents-mcpInstallieren Sie Abhängigkeiten:
npm installErstellen Sie
.envDatei: Kopieren Sie.env.examplenach.envund fügen Sie IhrenSERPAPI_API_KEYhinzu.Bauen:
npm run buildLokal ausführen:
npm startOder für die Entwicklung mit automatischem Rebuild:
npm run dev
Protokollierung
Protokolle werden als Standardfehler ausgegeben.
Die Protokollebene kann über die Umgebungsvariable
LOG_LEVELgesteuert werden (error,warn,info,http,verbose,debug,silly). Der Standardwert istinfo.Es wird versucht, eine Protokolldatei im Projektstammverzeichnis (
google-patents-server.log), im Home-Verzeichnis des Benutzers (~/.google-patents-server.log) oder/tmp/google-patents-server.logzu erstellen.
Lizenz
MIT-Lizenz (siehe LICENSE-Datei)
Available Tools
1 toolsearch_patentsC
Searches Google Patents using SerpApi. Allows filtering by date, inventor, assignee, country, language, status, type, and sorting.
| Name | Required | Description | Default |
|---|---|---|---|
| after | No | Minimum date filter (e.g., 'publication:20230101', 'filing:20220601'). Format: type:YYYYMMDD where type is 'priority', 'filing', or 'publication'. | |
| assignee | No | Filter by assignee names. Separate multiple names with a comma (,). | |
| before | No | Maximum date filter (e.g., 'publication:20231231', 'filing:20220101'). Format: type:YYYYMMDD where type is 'priority', 'filing', or 'publication'. | |
| country | No | Filter by country codes (e.g., 'US', 'WO,JP'). Separate multiple codes with a comma (,). | |
| inventor | No | Filter by inventor names. Separate multiple names with a comma (,). | |
| language | No | Filter by language (e.g., 'ENGLISH', 'JAPANESE,GERMAN'). Separate multiple languages with a comma (,). Supported: ENGLISH, GERMAN, CHINESE, FRENCH, SPANISH, ARABIC, JAPANESE, KOREAN, PORTUGUESE, RUSSIAN, ITALIAN, DUTCH, SWEDISH, FINNISH, NORWEGIAN, DANISH. | |
| num | No | Number of results per page (default: 10). **IMPORTANT: Must be 10 or greater (up to 100).** | |
| page | No | Page number for pagination (default: 1). | |
| q | Yes | Search query (required). Although optional in SerpApi docs, a non-empty query is practically needed. Use semicolon (;) to separate multiple terms. Advanced syntax like '(Coffee) OR (Tea);(A47J)' is supported. See 'About Google Patents' for details. | |
| scholar | No | Include Google Scholar results (default: false). | |
| sort | No | Sorting method. 'relevance' (default), 'new' (newest by filing/publication date), 'old' (oldest by filing/publication date). | relevance |
| status | No | Filter by patent status: 'GRANT' or 'APPLICATION'. | |
| type | No | Filter by patent type: 'PATENT' or 'DESIGN'. |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries full burden but offers minimal behavioral insight. It mentions filtering and sorting but omits critical details like rate limits, authentication needs, pagination behavior, error handling, or response format. For a search tool with 13 parameters, this is insufficient.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, efficient sentence that front-loads the core purpose and enumerates key capabilities without redundancy. Every word serves a purpose, making it highly concise and well-structured.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a complex search tool with 13 parameters, no annotations, and no output schema, the description is incomplete. It lacks behavioral context, usage guidelines, and details on results or errors, leaving significant gaps for agent understanding.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema fully documents all 13 parameters. The description adds no additional parameter semantics beyond listing filter types, which are already covered in schema descriptions. Baseline 3 is appropriate as the schema does the heavy lifting.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('Searches Google Patents') and resource ('patents'), and mentions the service provider ('using SerpApi'). It lists key filtering capabilities, making the purpose specific and actionable. However, without sibling tools, it cannot differentiate from alternatives, preventing a perfect score.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives, prerequisites, or typical use cases. It lists filtering options but lacks context on optimal scenarios or limitations, leaving the agent without usage direction.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
1 tool update
v1.0.0- First observed
search_patents
TDQS
Scored across 1 tool
With only one tool, there is no possibility of ambiguity or confusion between tools. The single tool has a clear, distinct purpose focused on searching patents.
The single tool name 'search_patents' follows a consistent verb_noun pattern. Since there is only one tool, naming consistency is inherently perfect with no deviations to assess.
A single tool for a patents server feels thin and under-scoped. While search is a core function, the domain suggests potential for additional operations like retrieving patent details, analyzing trends, or filtering by categories, making the count too low for the apparent scope.
The tool surface is severely incomplete for a patents domain. It only provides search functionality, with obvious gaps such as retrieving full patent documents, getting metadata like citations or legal status, or managing saved searches, which limits agent workflows and causes dead ends.
Maintenance
Related MCP Connectors
Official SerpApi MCP server for Google, Bing, and other search engines.
A Model Context Protocol server for Wix AI tools
MCP server for Google search results via SERP API
Google SERP, Maps, Shopping, News and AI-answer data for AI agents, from the Searlo search API.
Related MCP Servers
- AlicenseBqualityCmaintenanceA Model Context Protocol server that provides Google Jobs search capabilities through SerpAPI integration, featuring multi-language support, flexible search parameters, and rich job details.114 npm23MIT
- FlicenseCqualityCmaintenanceA Model Context Protocol server that enables LLMs to perform Google searches via the Serper API, allowing models to retrieve current information from the web.1339-
- AlicenseNot gradedqualityDmaintenanceA Model Context Protocol server that enables AI assistants to perform web searches using Google Search API, returning up to 20 search results in JSON format.2Apache 2.0
- AlicenseAqualityCmaintenanceMCP server for patent search and prior art discovery powered by Google Patents public dataset on BigQuery. Supports searching patents, fetching full patent details with CPC codes and citations, and retrieving legal claims text.36MIT