Claude Web Search MCP Server
Claude Web 搜索 MCP 服务器
该 MCP(模型上下文协议)服务器使用 Claude API 提供网络搜索功能。它允许 LLM 通过标准化接口访问网络上的最新信息。
特征
使用 Claude 的网络搜索 API 的网络搜索工具
支持域过滤(允许和阻止域)
可配置每次搜索的最大结果数
从 Claude Desktop 配置文件自动配置
Related MCP server: Google Search MCP
先决条件
Node.js 18 或更高版本
启用了网络搜索的 Anthropic API 密钥
用于测试的 Claude 桌面应用程序
安装和设置
克隆存储库:
git clone https://github.com/Doriandarko/claude-search-mcp.git cd claude-search-mcp安装依赖项:
npm install构建服务器:
npm run build这将编译 TypeScript 代码并使服务器可执行。
链接服务器以实现全局访问:
npm link这使得
mcp-server-claude-search命令在整个系统范围内可用,从而允许 Claude Desktop 应用程序找到它。
使用 Claude 桌面应用程序运行服务器
一旦服务器安装并链接,如果配置正确,Claude Desktop 应用程序就可以自动管理它。
**配置 Claude 桌面应用:**打开 Claude 桌面应用的 MCP 服务器配置文件(通常为
claude_desktop_config.json)。添加或更新此服务器的条目:{ "mcpServers": { // ... other servers ... "claude-search": { "command": "mcp-server-claude-search", "env": { "ANTHROPIC_API_KEY": "YOUR_ANTHROPIC_API_KEY_HERE" } } // ... other servers ... } }将
"YOUR_ANTHROPIC_API_KEY_HERE"替换为你实际的 Anthropic API 密钥。如果env变量未在此处设置,服务器也会尝试从~/code/claude-search-mcp/claude_desktop_config.json读取此密钥,但建议在主配置中为每个服务器分别定义此密钥。**启动 Claude 桌面应用程序:**启动(或重启)您的 Claude 桌面应用程序。现在它应该能够在需要时找到并启动
mcp-server-claude-search。**使用网络搜索:**您现在可以在与 Claude 的对话中使用网络搜索功能。
手动服务器执行(用于测试/开发)
如果您想手动运行服务器以进行测试或开发目的(在 Claude Desktop 应用程序管理之外):
使用全局链接命令:
mcp-server-claude-search直接使用 tsx(用于自动重启开发):
npm run dev直接运行编译后的代码:
npm start
Web 搜索工具参数
当 LLM 调用网络搜索工具时,它支持以下参数:
query(必需):搜索查询字符串。maxResults(可选):返回的最大搜索结果数(默认值:5)。allowedDomains(可选):要包含在搜索结果中的域数组(例如,["example.com", "wikipedia.org"])。blockedDomains(可选):从搜索结果中排除的域数组。
执照
麻省理工学院
Available Tools
1 toolweb_searchA
Search the web for real-time information about any topic. Use this tool when you need up-to-date information that might not be available in your training data, or when you need to verify current facts.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | The search query to look up on the web | |
| maxResults | No | Maximum number of search results to return (default: 5) | |
| allowedDomains | No | Only include results from these domains | |
| blockedDomains | No | Never include results from these domains |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It mentions the tool's purpose and usage context but lacks details on behavioral traits such as rate limits, authentication needs, or specific output format. The description does not contradict any annotations, but it could be more informative about operational constraints.
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 front-loaded and concise, consisting of two sentences that directly address purpose and usage guidelines without unnecessary details. Every sentence earns its place by providing essential information efficiently.
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?
Given the tool's complexity (web search with 4 parameters) and no output schema, the description is adequate but could be more complete. It covers purpose and usage well but lacks details on behavioral aspects like result format or limitations. With no annotations, it should ideally provide more context to fully guide the agent.
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?
The input schema has 100% description coverage, so the schema already documents all parameters well. The description does not add any parameter-specific information beyond what the schema provides, such as examples or additional context. This meets the baseline for high schema coverage, but no extra value is added.
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 tool's purpose with specific verbs ('search the web') and resources ('real-time information about any topic'). It distinguishes the tool's function from potential alternatives by emphasizing real-time, up-to-date information that might not be in training data, which is essential for a web search tool.
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 explicitly states when to use this tool: 'when you need up-to-date information that might not be available in your training data, or when you need to verify current facts.' This provides clear context and guidance for the agent, even though no sibling tools are listed, making it comprehensive for decision-making.
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
- First observed
web_search
TDQS
Scored across 1 tool
With only one tool, there is no possibility of confusion or overlap between tools, as there are no other tools to compare it against. The single tool's purpose is clearly defined and distinct by default.
The single tool name 'web_search' follows a consistent verb_noun pattern, and since there are no other tools, there is no inconsistency to evaluate. The naming is straightforward and appropriate for its function.
A single tool is generally too few for most server purposes, as it limits functionality and can feel thin or incomplete. For a web search server, while the core function is covered, additional tools like advanced search filters or result management might be expected to enhance utility.
The server provides a basic search function, which covers the essential need for real-time information retrieval. However, there are notable gaps, such as lack of tools for refining searches, handling pagination, or accessing specific types of content, which could limit agent effectiveness in complex scenarios.
Maintenance
Related MCP Connectors
Provides AI assistants with access to Seltz's powerful Web Search capabilities.
Web search and page-reading for AI agents. One-click OAuth connect, or a Caesar API key.
Web search, browser automation, scraping, crawling and CAPTCHA solving for AI agents.
The best web search for your AI Agent
Related MCP Servers
- AlicenseBqualityCmaintenanceFacilitates web search capabilities using Perplexity's API, allowing users to retrieve search results through Claude's interface.18MIT
- AlicenseBqualityDmaintenanceEnables Large Language Models to perform real-time web searches using Google Custom Search API. Integrates with Claude Desktop to retrieve current information from the internet.1MIT
- AlicenseNot gradedqualityDmaintenanceExposes Perplexity AI's search capabilities to Claude, enabling real-time web search and information retrieval within the assistant. The project is currently in active development with plans to support Perplexity Spaces and multi-source data synthesis.Apache 2.0
- AlicenseNot gradedqualityDmaintenanceEnables web search via Brave Search API with freshness filters, full page content fetching, and usage tracking. Designed for Claude integration.399 npmGPL 3.0