Integration Guide

MCP Server Integration Guide

Connect KhaleejiAPI to MCP-compatible AI clients like VS Code Copilot and Claude Desktop. This guide covers setup, required configuration, and a minimal working tool-call flow.

1) Install and run

Run the MCP server with your KhaleejiAPI key in environment variables.

bash
npx @khaleejiapi/mcp-server

2) Add client configuration

Add one of the following configs, depending on your MCP client.

VS Code / GitHub Copilot

Save in .vscode/mcp.json.

json
{
  "servers": {
    "khaleejiapi": {
      "command": "npx",
      "args": ["@khaleejiapi/mcp-server"],
      "env": {
        "KHALEEJI_API_KEY": "your_api_key_here"
      }
    }
  }
}

Claude Desktop

Save in claude_desktop_config.json.

json
{
  "mcpServers": {
    "khaleejiapi": {
      "command": "npx",
      "args": ["@khaleejiapi/mcp-server"],
      "env": {
        "KHALEEJI_API_KEY": "your_api_key_here"
      }
    }
  }
}

3) Minimal working example

Ask your assistant a concrete prompt that triggers a KhaleejiAPI tool.

text
User prompt:
"Validate [email protected] and summarize if it is safe for onboarding emails."
 
Expected tool call:
validate_email({ "email": "[email protected]" })

If the setup is correct, your client should call an MCP tool such as validate_email and return structured output.

Required environment variables

bash
KHALEEJI_API_KEY=your_api_key_here
KHALEEJI_BASE_URL=https://khaleejiapi.dev/api/v1