MCP server
Give Claude Desktop real-time web access
Connect webclaw over MCP and Claude can read any webpage.
Claude Desktop supports the Model Context Protocol natively. Add webclaw over MCP and Claude gets 14 web tools immediately: scrape, search, crawl, map, batch, extract, summarize, diff, brand, research, lead, lead_batch, list_extractors, and vertical_scrape. No middleware, no custom code.
Setup
claude_desktop_config.json
{
"mcpServers": {
"webclaw": {
"command": "npx",
"args": ["-y", "@webclaw/mcp"],
"env": {
"WEBCLAW_API_KEY": "wc_..."
}
}
}
}Why webclaw for Claude Desktop
- Native MCP server, no wrappers
- 14 tools exposed in one config
- Handles bot-protected sites automatically
- Fewer 403s on protected docs and forums
Common use cases
- Ask Claude to research topics across the web
- Extract structured data from URLs
- Summarize long articles and documentation
- Build personal knowledge bases from web content
Frequently asked questions.
How do I add webclaw to Claude Desktop?
Install webclaw via cargo install webclaw or download a release binary. Then add the MCP config shown above to claude_desktop_config.json and restart Claude Desktop.
What tools become available to Claude?
14 tools: scrape, search, crawl, map, batch, extract, summarize, diff, brand, research, lead, lead_batch, list_extractors, and vertical_scrape — covering single-page and multi-URL extraction, web search, crawling, structured data, change detection, brand signals, deep research, company enrichment, and site-specific extractors.
Other integrations