← All articles

Claude MCP Setup Guide

June 15, 2026·5 min read·MCPForge

Claude MCP Setup Guide


This guide walks through connecting Claude Desktop to an MCP server. Once set up, Claude can call your API directly from the conversation.


Prerequisites


  • Claude Desktop installed (available at claude.ai/download)
  • An MCP server endpoint URL

Step 1: Find Your Config File


Claude Desktop reads MCP server configuration from a JSON file. The location depends on your operating system:


  • **macOS**: `~/Library/Application Support/Claude/claude_desktop_config.json`
  • **Windows**: `%APPDATA%\Claude\claude_desktop_config.json`

If the file doesn't exist, create it.


Step 2: Add Your MCP Server


Open the config file and add your MCP server:


{

"mcpServers": {

Want to analyze your API security?

Import your OpenAPI spec and generate a Security Report automatically.

Generate Security Report →

"my-api": {

"url": "https://www.mcpforge.tech/api/servers/my-api-slug",

"transport": "http"

}

}

}


If you're using MCPForge, the config snippet is generated automatically. Copy it from your dashboard and paste it into the config file.


Step 3: Restart Claude Desktop


After saving the config file, restart Claude Desktop. It will connect to your MCP server on startup and load the available tools.


Step 4: Verify the Connection


In Claude Desktop, open a new conversation and ask:


> "What tools do you have available?"


Claude should list the tools from your MCP server. If it doesn't, check:


  • The config file is valid JSON (no trailing commas, correct quotes)
  • The URL is correct and accessible
  • Claude Desktop was fully restarted (not just the window)

Step 5: Test a Tool Call


Try a simple read operation first:


> "List my customers" or "Show me the available products"


If the tool call succeeds, you'll see the results in Claude's response. If it fails, Claude will tell you what went wrong.


Common Issues


**Tool names not showing up**: The MCP server may not be running or the URL may be incorrect. Check that the endpoint returns a valid response.


**Authentication errors**: Your API requires credentials. Add them to the MCP server's credential vault rather than the config file.


**Tool call fails with "disabled" error**: The tool has been disabled in the MCP server's permission settings. Enable it in the dashboard.


Next Steps


Once your MCP server is connected:


1. Review which tools are exposed — disable any you don't want Claude to access

2. Add API credentials to the server-side vault

3. Enable approval workflows for destructive operations

4. Check the audit log to see what Claude is calling


MCPForge provides all of these features through a dashboard — no code required.

Check your MCP security posture

Generate a Security Score, detect risky tools, and review permissions before exposing APIs to AI agents.

Try MCPForge Free →

Related Articles

What Is Model Context Protocol (MCP)?

OpenAPI to MCP: Complete Guide

How to Connect Claude to Any API Using MCP

Coming soon

GitHub MCP Server Explained

Coming soon