AgentQL Data Extractor MCP Server

Browser AutomationTypeScript

Extract structured data from web pages using natural language prompts

Available Tools

extract-web-data

Extract structured data from a given URL using a natural language prompt to describe the data and fields to extract

urlprompt

AgentQL Data Extractor provides powerful web data extraction capabilities through a Model Context Protocol (MCP) server. It allows AI assistants to extract structured information from websites based on natural language descriptions of the data you need. This integration enables your AI assistant to gather real-time information from the web without requiring complex coding or web scraping knowledge.

Overview

AgentQL Data Extractor is a Model Context Protocol (MCP) server that enables AI assistants to extract structured data from web pages. By simply describing what information you need in natural language, the tool can navigate websites and return organized data in your preferred format.

Prerequisites

Before installing AgentQL Data Extractor, you'll need:

Node.js and npm installed on your system
An API key from the AgentQL Dev Portal (https://dev.agentql.com/)

Installation

You can install the AgentQL Data Extractor globally via npm:

npm install -g agentql-mcp

Configuration

Claude Desktop

To configure AgentQL Data Extractor in Claude Desktop:

Open Claude Desktop Settings (⌘ + ,)
Navigate to the "Developer" section in the sidebar
Click "Edit Config" to open the claude_desktop_config.json file
Add the AgentQL server configuration to the mcpServers section
Restart Claude Desktop

Cursor

To set up AgentQL Data Extractor in Cursor:

Open Cursor Settings
Go to "MCP > MCP Servers"
Click "+ Add new MCP Server"
Enter the required configuration details:
- Name: "agentql" (or your preferred name)
- Type: "command"
- Command: env AGENTQL_API_KEY=YOUR_API_KEY npx -y agentql-mcp

Windsurf

To configure AgentQL Data Extractor in Windsurf:

Open "Windsurf: MCP Configuration Panel"
Click "Add custom server+"
Alternatively, edit ~/.codeium/windsurf/mcp_config.json directly
Add the AgentQL server configuration to the mcpServers section

Usage

Once configured, you can ask your AI assistant to extract data from websites. For example:

"Extract the list of videos from the YouTube search results for 'agentql', including title, author name, view count, and URL. Format as a markdown table."

If your assistant doesn't automatically use the AgentQL tool, try adding hints like "use tools" or "use agentql tool" to your prompt.

Troubleshooting

If you encounter issues:

Verify your API key is correct
Ensure the MCP server is properly configured in your AI assistant
Check that you've restarted your AI assistant after configuration
Try explicitly instructing your assistant to use the AgentQL tool

For additional support, visit the AgentQL documentation.