Overview
This guide explains how customers can connect to Limble's Official MCP Server. It covers authentication and configuration for popular MCP clients: Cursor, Claude Desktop, Claude Code, GitHub Copilot, and Microsoft Copilot Studio.
Table of Contents
Prerequisites
Download and install Node.js: https://nodejs.org/en/download
Hosted MCP endpoint URL, for example:
<https://mcp.limblecmms.com/mcp)>Limble API credentials:
Limble-Client-ID: <id>andLimble-Client-Secret: <secret>headersLearn more about managing API keys in this article.
The server handles authentication injection to Limble and applies a consistent User-Agent automatically.
Step 1: Configure Your MCP Client
Select your client below (Cursor IDE, Claude Desktop, Claude Code, GitHub Copilot, or Microsoft Copilot Studio) and follow the configuration steps.
Cursor IDE
Cursor IDE
Edit ~/.cursor/mcp.json and add an entry that uses the hosted endpoint via mcp-remote:
{
"mcpServers": {
"limble-mcp": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.limblecmms.com/mcp",
"--header",
"Limble-Client-ID: YOUR_CLIENT_ID",
"--header",
"Limble-Client-Secret: YOUR_CLIENT_SECRET"
]
}
}
}
After saving, restart Cursor.
Claude Desktop
Claude Desktop
Open Claude Desktop → Settings → Developers → Edit Config and add a hosted entry using split headers:
{
"mcpServers": {
"limble-mcp": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.limblecmms.com/mcp",
"--header",
"Limble-Client-ID: YOUR_CLIENT_ID",
"--header",
"Limble-Client-Secret: YOUR_CLIENT_SECRET"
]
}
}
}After saving, restart Claude Desktop.
Claude Code
Claude Code
Edit or create /YOUR_PROJECT_DIRECTORY/.mcp.json and add an entry that uses the hosted endpoint via mcp-remote:
{
"mcpServers": {
"limble-mcp": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.limblecmms.com/mcp",
"--header",
"Limble-Client-ID: YOUR_CLIENT_ID",
"--header",
"Limble-Client-Secret: YOUR_CLIENT_SECRET"
]
}
}
}
After saving, restart Claude Code.
GitHub Copilot
GitHub Copilot
GitHub Copilot runs inside code editors such as Visual Studio Code, Visual Studio, and JetBrains IDEs. MCP servers are configured per-project using a .vscode/mcp.json file. For full setup details, reference the GitHub Copilot MCP documentation.
Create or edit .vscode/mcp.json in your project directory and add the following:
{
"inputs": [
{
"type": "promptString",
"description": "Limble Client ID",
"id": "limbleClientId",
"default": "YOUR_CLIENT_ID",
"password": false
},
{
"type": "promptString",
"description": "Limble Client Secret",
"id": "limbleClientSecret",
"default": "YOUR_CLIENT_SECRET",
"password": true
}
],
"servers": {
"limble-mcp": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.limblecmms.com/mcp",
"--header",
"Limble-Client-ID: ${input:limbleClientId}",
"--header",
"Limble-Client-Secret: ${input:limbleClientSecret}"
]
}
}
}
The inputs block tells GitHub Copilot to prompt you for your credentials at startup rather than storing them in the file. The password: true flag on the secret field masks your input while typing. After saving, reload or restart your IDE. GitHub Copilot will prompt for your Client ID and Client Secret the first time the MCP server is used.
Microsoft Copilot Studio
Microsoft Copilot Studio
Microsoft Copilot Studio is a no-code platform for building and extending AI agents. MCP servers are added through the Copilot Studio UI, and no config file is required. For full setup details, see the Microsoft Copilot Studio MCP documentation.
Note on authentication: Copilot Studio connection setup for Limble will require HTTP Basic authentication, which combines your Client ID and Client Secret into a single base64-encoded |
Generate Your Authorization Value
Before starting, encode your credentials. One option is to use a terminal shell command to accomplish this task.
For example, replace YOUR_CLIENT_ID and YOUR_CLIENT_SECRET with your actual values and run the following in a terminal:
echo -n "YOUR_CLIENT_ID:YOUR_CLIENT_SECRET" | base64
Copy the output. You will need it in the wizard below.
Steps
Log into Copilot Studio.
Create a new agent, or open an existing one and skip to step 4.
Complete the agent creation wizard and open the agent.
Select Add Tools from the agent panel.
Choose Model Context Protocol (MCP) from the tool type list.
If you do not see it, select Add first, then choose Model Context Protocol.
Fill in the MCP connection wizard:
Field | Value |
Server URL |
|
Header Name |
|
Key Value |
|
7. Click Add.
Verify the connection
Once the server is connected, test the agent with a prompt such as:
Limble list_mcp_tools
A successful response lists the available Limble MCP tools and confirms the connection is working.
Step 2: Verify the Connection
In your MCP client, try any of the following:
help()— overview and discoveryget_current_customer_info()— verifies authenticationget_tasks({"limit": 1})— quick data fetchlist_mcp_tools({"format": "summary"})— available tools
Troubleshooting
403 (Unauthorized): Check your credentials or header format.
Endpoint unreachable: Verify the domain, path (
/mcp).Config syntax: Ensure JSON is valid (commas, quotes).
Client caching: Restart Cursor/Claude after changes.
Related Articles
Still don’t see what you’re looking for? Check out our YouTube channel for more tips and tricks! You can also reach out to us anytime at support@limblecmms.com.