brightdata-mcp  by brightdata

MCP server for real-time web data access in AI agents

Created 5 months ago
1,321 stars

Top 30.3% on SourcePulse

GitHubView on GitHub
Project Summary

This project provides a Model Context Protocol (MCP) server for AI agents and LLMs to access real-time web data, overcoming common blocking mechanisms. It targets developers building AI-powered applications that require up-to-date information from the web, offering seamless integration with MCP-compatible clients and bypassing geo-restrictions.

How It Works

The server leverages Bright Data's infrastructure, including its Web Unlocker and optional Browser API, to facilitate web scraping. It acts as a proxy, allowing AI agents to issue commands for searching, navigating, and extracting data from websites. This approach bypasses geo-restrictions and bot detection, providing reliable access to public web data.

Quick Start & Requirements

Highlighted Details

  • Real-time web access for AI agents.
  • Bypasses geo-restrictions and bot detection.
  • Supports remote browser automation.
  • Integrates with MCP-compatible AI assistants like Claude Desktop, Cursor, and Windsurf.

Maintenance & Community

  • Maintained by Bright Data.
  • Contributions welcome via issues and pull requests.
  • Coding style guidelines provided.
  • Support available through Bright Data or GitHub issues.

Licensing & Compatibility

  • License: Not explicitly stated in the README.
  • Compatibility: Designed for MCP-compatible clients. Commercial use is dependent on Bright Data's terms of service.

Limitations & Caveats

Scraped web content should be treated as untrusted and validated before use in LLM prompts to prevent prompt injection. Timeouts may need adjustment for slow-loading websites. The npx command might require path configuration if not found by the system.

Health Check
Last Commit

17 hours ago

Responsiveness

1 day

Pull Requests (30d)
8
Issues (30d)
8
Star History
228 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.