Web Tools MCP Server

A TypeScript MCP (Model Context Protocol) server that provides web search and intelligent research capabilities through Claude CLI integration.

Features

• research_agent: Intelligent research agent that answers questions by searching and analyzing web sources

• web_search: Search the web for real-time information with domain filtering support

• web_fetch: Fetch content from specific URLs

• Uses Claude CLI backend for actual web operations

• Full TypeScript implementation with proper error handling

• Compatible with Claude Desktop and other MCP clients

Related MCP server: MCP Deep Web Research Server

Installation

Claude Desktop Integration

Add the following configuration to your Claude Desktop config file:

Config File Locations:

• Windows: %APPDATA%\Claude\claude_desktop_config.json

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

• Linux: ~/.config/Claude/claude_desktop_config.json

Basic Configuration

Custom Claude CLI Path

If you need to specify a custom Claude CLI path:

Tool Usage

research_agent

Intelligent research agent that automatically searches, fetches, and analyzes web sources to answer research questions.

Parameters:

• objectives (required): Array of research questions to answer

• starting_query (required): Initial search query to begin research

• max_calls (optional): Maximum research iterations (default: 5)

• allowed_domains (optional): Array of domains to include in results

• blocked_domains (optional): Array of domains to exclude from results

Features:

• Automatically searches the web and fetches top 3 most relevant sources

• Extracts comprehensive quotes from web content

• Continues research until objectives are met or max iterations reached

• Provides organized results with source citations

web_search

Search the web for information with optional domain filtering.

Parameters:

• query (required): The search query string

• allowed_domains (optional): Array of domains to include in results

• blocked_domains (optional): Array of domains to exclude from results

web_fetch

Fetch and extract content from a specific URL.

Parameters:

• url (required): The URL to fetch content from

Usage Examples

Basic Research

Research with Domain Filtering

Basic Web Search

How Research Agent Works

1. Search: Performs web searches for the research objectives

2. Fetch: Automatically retrieves content from the top 3 most relevant sources

3. Analyze: Extracts comprehensive quotes that answer research questions

4. Continue: Iterates until objectives are met or max iterations reached

Requirements

• Node.js >= 18.0.0

• Claude CLI installed and accessible

• Compatible with MCP client applications

Development