- A powerful web scraping MCP server with 1.6k+ stars on GitHub
- Features:
- Web scraping with JavaScript rendering
- URL discovery and crawling
- Web search with content extraction
- Batch processing with rate limiting
- Mobile/Desktop viewport support
- Smart content filtering
- Installation:
npx -y firecrawl-mcp - Requires API key from firecrawl.dev
- Official Brave Search API integration for MCP
- Provides web and local search capabilities
- Features:
- Web search results
- News articles
- Image search
- Local results
- Installation:
env BRAVE_API_KEY=[your-key] npx -y @modelcontextprotocol/server-brave-search - Requires Brave Search API key
- Vector search engine MCP server with semantic memory capabilities
- Features:
- Store and retrieve information with vector search
- Semantic code snippet storage
- Metadata support
- SSE transport support for remote connections
- Installation: Requires running Qdrant instance
- GitHub: 260+ stars, Apache 2.0 license
- Browser automation and control MCP server
- Features:
- Automated UI testing
- Web scraping and research
- Complex web task automation
- Documentation checking
- Integration testing support
- Installation: Available via Smithery
- Best suited for:
- Frontend development workflows
- Web application testing
- Research and documentation tasks
- Browser-based automation
- Primary registry for MCP servers
- Features:
- Searchable directory of MCP servers
- Quality ratings and reviews
- Easy installation commands
- Integration with Cursor IDE
- Website: smithery.ai
- Official Cursor IDE MCP server directory
- Features:
- Curated list of tested MCP servers
- Installation instructions
- Compatibility information
- User ratings and reviews
- Website: cursor.directory/mcp
- Community-driven MCP server directory
- Features:
- Open-source server listings
- Documentation and guides
- Integration examples
- Community contributions
- Focus on developer tools and utilities
-
Server Installation:
- Use stdio/local setup for personal use
- Use SSE transport for team/remote setups
- Always secure API keys properly
-
Configuration:
- Add servers one at a time
- Verify status indicators
- Test functionality before production use
-
Usage:
- MCP works in Composer and Agent modes
- Explicitly instruct AI to use MCP tools
- Accept tool usage prompts when they appear
-
Troubleshooting:
- Check server status indicators
- Verify API key formatting
- Restart Cursor if tools don't appear
- Update to latest Cursor version