Overview
Understanding Model Context Protocol for AI development
Model Context Protocol (MCP) enables AI assistants to connect with external tools and data sources, transforming them from text processors into active development partners. The codebase has baked-in support for Context7, automatically using it for BetterAuth and UseAutumn documentation when the MCP server is available.
Enhanced AI Capabilities
MCP servers provide AI assistants with real-time access to documentation, UI component generation, and external tools. StarterApp automatically uses Context7 for BetterAuth and UseAutumn documentation when available.
What MCP Provides
MCP creates standardized connections between AI assistants and development tools. Instead of copying and pasting between chat interfaces and terminals, AI assistants gain direct access to documentation, file systems, and command execution. This integration eliminates context switching while ensuring AI suggestions remain current and executable.
Quick Start Installation
Each AI tool requires specific configuration to enable MCP servers. The following guides cover installation for common development environments.
Claude Code
Configure MCP servers through command line:
# Install Context7 for documentation access
claude mcp add --transport http context7 https://mcp.context7.com/mcp --header "CONTEXT7_API_KEY: YOUR_API_KEY"
# Configuration stored in claude_mcp.json
Claude Code automatically detects available MCP servers and provides tool suggestions during conversations.
GitHub Copilot (VS Code)
Install MCP servers through VS Code extensions:
- Open Extensions view (⇧⌘X or Ctrl+Shift+X)
- Search for
@mcp
to find available servers - Install desired MCP servers
- Run command:
MCP: Show Installed Servers
to verify
Configuration example:
{
"github.copilot.chat.mcp.enabled": true,
"chat.mcp.discovery.enabled": true
}
Cursor
Create MCP configuration in project root:
# Create .cursor directory if not exists
mkdir .cursor
# Create mcp.json configuration
cat > .cursor/mcp.json << EOF
{
"mcpServers": {
"context7": {
"serverUrl": "https://mcp.context7.com/mcp",
"headers": {
"CONTEXT7_API_KEY": "YOUR_API_KEY"
}
}
}
}
EOF
Cursor loads MCP servers on startup and makes tools available through #
commands.
Windsurf
Add servers to Windsurf configuration:
{
"mcpServers": {
"context7": {
"serverUrl": "https://mcp.context7.com/mcp",
"headers": {
"CONTEXT7_API_KEY": "YOUR_API_KEY"
}
},
"magic": {
"command": "npx",
"args": ["-y", "@magic/mcp-server"]
}
}
}
Windsurf integrates MCP tools into the chat interface automatically.
Supported MCP Servers
Context7
Real-time documentation access for UseAutumn, BetterAuth, Convex, and thousands of libraries.
Magic MCP
UI component generation from natural language with production-ready React code.
Related Documentation
AI Workflow Overview
Understanding AI-assisted development patterns and tooling.
Commands
AI command reference for development workflows.
Templates
Production-ready code patterns for consistent AI generation.