Installation
Prerequisites
Before installing, you need a Gridleaf API key:
- Sign up at gridleaf.org/auth/signup (free, no credit card required)
- Go to your Dashboard and create an API key
- Copy your key (format:
gk_...) — you'll need it below
All platforms connect to the same unified gateway at https://mcp.gridleaf.org/mcp, which gives you access to all 50+ tools through a single connection.
Claude Code (CLI) — Recommended
Run this single command in your terminal:
claude mcp add gridleaf --transport http --url "https://mcp.gridleaf.org/mcp" --header "X-API-Key: YOUR_API_KEY"Replace YOUR_API_KEY with your actual API key. That's it — all tools are immediately available.
Claude Desktop
- Open your Claude Desktop configuration file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- Add the Gridleaf MCP server configuration:
{
"mcpServers": {
"gridleaf": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.gridleaf.org/sse",
"--header",
"X-API-Key: YOUR_API_KEY"
]
}
}
}- Replace
YOUR_API_KEYwith your actual API key - Restart Claude Desktop — all 50+ tools will be available
Claude.ai (Web)
- Go to Claude.ai Settings → Integrations
- Click "Add Integration" and select "MCP Server"
- Enter the server URL:
https://mcp.gridleaf.org/mcp - Add your API key when prompted
Cursor
- Open Cursor Settings (Cmd/Ctrl + ,)
- Navigate to the MCP section
- Add a new MCP server with the following SSE URL:
https://mcp.gridleaf.org/sse- Set the header
X-API-Keyto your API key - The tools will be available in your next session
Windsurf / Other MCP Clients
Any MCP-compatible client can connect using:
- Streamable HTTP endpoint:
https://mcp.gridleaf.org/mcp - SSE endpoint:
https://mcp.gridleaf.org/sse - Authentication header:
X-API-Key: YOUR_API_KEY
Verify Your Connection
After setup, try asking your AI assistant:
- "What tools do you have from Gridleaf?"
- "Show me the carbon intensity ranking for US ISOs"
If you see results, you're all set! If not, double-check your API key and restart your AI assistant.