Use with VS Code
Windsurf has first-class MCP support. VS Code itself doesn't, but Continue, Cline, and Copilot Chat all do — same Crawlbase config, four places to drop it.
Windsurf (officially supported)
Windsurf is an officially supported client for Crawlbase MCP. Configuration:
- Open Windsurf → File → Preferences → Windsurf Settings → General → MCP Servers → Manage MCPs → View raw config.
- Add the Crawlbase entry to
mcp_config.json:
{
"mcpServers": {
"crawlbase": {
"type": "stdio",
"command": "npx",
"args": ["@crawlbase/mcp@latest"],
"env": {
"CRAWLBASE_TOKEN": "YOUR_TOKEN",
"CRAWLBASE_JS_TOKEN": "YOUR_JS_TOKEN"
}
}
}
}Save and restart Windsurf. The Crawlbase tools appear in the MCP panel.
VS Code (via assistant extensions)
VS Code itself doesn't ship MCP support, but the popular AI assistants for it do. Pick whichever you already use.
Continue
Continue is a popular open-source AI assistant for VS Code. It supports MCP servers in its YAML config:
mcpServers:
- name: crawlbase
type: stdio
command: npx
args:
- "@crawlbase/mcp@latest"
env:
CRAWLBASE_TOKEN: YOUR_TOKEN
CRAWLBASE_JS_TOKEN: YOUR_JS_TOKENReload Continue and the tools appear in the agent panel.
Cline
Cline (formerly Claude Dev) has a settings UI for MCP servers. Open Cline → Settings → MCP Servers → Add, and paste:
{
"crawlbase": {
"type": "stdio",
"command": "npx",
"args": ["@crawlbase/mcp@latest"],
"env": {
"CRAWLBASE_TOKEN": "YOUR_TOKEN",
"CRAWLBASE_JS_TOKEN": "YOUR_JS_TOKEN"
}
}
}GitHub Copilot Chat
Copilot Chat has MCP support. Open VS Code Settings, search for "MCP", enable the integration, then add via the same JSON shape used above to ~/.config/Code/User/mcp.json.
All four options use identical Crawlbase config. Windsurf is the official upstream-supported client; the three VS Code options work just as well — Continue is most extensible, Cline is most agentic, Copilot is most native.
Usage
Once configured, the workflow is the same as in Cursor or Claude Desktop:
Crawl https://docs.example.com/openapi.json and generate a typed
client for the /users endpoints in TypeScript.The assistant calls crawl to fetch the spec, parses it, and edits files in your workspace.

