MCP Service Configuration
Configure AgileBuilder MCP service
Overview
This article covers how to start the AgileBuilder MCP service and configure it in various AI tools.
Starting the MCP Service
Basic Start
agilebuilder mcp start
By default, the service listens on http://localhost:3457.
Custom Port
agilebuilder mcp start --port 8080
Run in Background
agilebuilder mcp start --daemon
Check Status
agilebuilder mcp status
Configure AI Tools
Cursor
- Open Cursor settings
- Go to「MCP Servers」configuration
- Add a new server:
{
"mcp": {
"servers": {
"agilebuilder": {
"url": "http://localhost:3457"
}
}
}
}
Claude Code
Configure via command line:
# Method 1: Command line argument
claude code --mcp agilebuilder=http://localhost:3457
# Method 2: Configuration file
# Add to ~/.claude_desktop_config.json
{
"mcpServers": {
"agilebuilder": {
"url": "http://localhost:3457"
}
}
}
Windsurf
- Open Windsurf settings
- Go to「Extensions」or「MCP」configuration
- Add the AgileBuilder MCP server address
VS Code Copilot
Note: VS Code Copilot's MCP support depends on official updates. Please check for the latest developments.
Verify Configuration
After configuration, you can verify by asking:
"List all available AgileBuilder templates"
The AI should return the template list.
Troubleshooting Connection Issues
Service Not Running
# Check service status
agilebuilder mcp status
# Restart if not running
agilebuilder mcp start
Port Already in Use
# Use a different port
agilebuilder mcp start --port 8080
# Update the configuration in your AI tool
Network Issues
- Ensure the MCP service and AI tool can access the same network
- Check firewall settings
- For local services, use localhost
Security Considerations
Local Services
By default, the MCP service only allows local access. Do not expose the service port to public networks.
Authentication Configuration
For remote access, you can configure authentication:
agilebuilder mcp start --auth-token YOUR_TOKEN
Then configure in your AI tool:
{
"mcp": {
"servers": {
"agilebuilder": {
"url": "http://localhost:3457",
"headers": {
"Authorization": "Bearer YOUR_TOKEN"
}
}
}
}
}
Next Steps
- AI Workflow - Configure AI workflows
- AI Use Cases - Practical examples
- AI Tool Compatibility - Tool support details