How to set up and configure the Linkbreakers MCP server

Complete guide to installing, configuring, and using the Linkbreakers Model Context Protocol (MCP) server to connect AI assistants like Claude to your Linkbreakers workspace.

Developer
8 min read
By Laurent Schaffner
Updated February 5, 2026

Short answer

The Linkbreakers MCP server enables AI assistants (like Claude Desktop, Continue, and other MCP-compatible tools) to interact with your Linkbreakers workspace using natural language. It acts as a bridge between MCP clients and the Linkbreakers REST API, allowing AI to create links, analyze data, manage workflows, and more.

One-line install

Linkbreakers’ step-by-step setup makes it easy to add the MCP server in seconds.

curl -fsSL https://mcp.linkbreakers.com/install.sh | bash

Quick summary

  • MCP (Model Context Protocol) is an open standard that lets AI assistants connect to external tools and data
  • The Linkbreakers MCP server runs at https://mcp.linkbreakers.com
  • Setup takes under 5 minutes - just create a workspace token and configure your AI client
  • No installation required - use the hosted server
  • Supports 35+ curated operations across links, workflows, analytics, QR codes, and more

What is MCP and why use it?

Model Context Protocol (MCP) is an open protocol developed by Anthropic that standardizes how AI assistants connect to external data sources and tools. Instead of manually calling APIs or copying data, you can simply ask your AI assistant to:

"Create a shortened link for https://example.com with the slug 'demo'"

"Show me my workspace analytics for the last 30 days"

"List all links created this week"

"Generate a QR code for my latest campaign"

The MCP server translates these natural language requests into Linkbreakers API calls, handles authentication, and returns results in a format AI assistants understand.

Benefits

  • Natural language interface - No need to write API calls manually
  • Faster workflows - Ask questions and get answers instantly
  • Automated tasks - AI can execute multi-step operations
  • No code required - Non-technical users can access data programmatically

Prerequisites

Before you begin, you'll need:

  1. A Linkbreakers workspace with an active account
  2. An MCP-compatible client such as: Claude Desktop (macOS/Windows), Continue (VSCode extension), or any MCP-compatible AI assistant
  3. A workspace token (we'll create this in the next step)

Step 1: Create a workspace token

Workspace tokens are long-lived API credentials that authenticate your MCP client with Linkbreakers.

  1. Log in to your Linkbreakers account
  2. Navigate to Settings → API Tokens
  3. Click "Create Token"
  4. Enter a descriptive name like "Claude Desktop - MacBook"
  5. Click "Create"
  6. Copy the token immediately - it will only be shown once!

Step 2: Configure your MCP client

Configuration varies by client. Choose your AI assistant below:

Claude Desktop

  1. Locate your configuration file:

    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
    • Windows: %APPDATA%\Claude\claude_desktop_config.json
  2. Open the file in a text editor (create it if it doesn't exist)

  3. Add the Linkbreakers MCP server:

JSON
{
  "mcpServers": {
    "linkbreakers": {
      "url": "https://mcp.linkbreakers.com",
      "headers": {
        "Authorization": "Bearer YOUR_WORKSPACE_TOKEN_HERE"
      }
    }
  }
}
  1. Replace YOUR_WORKSPACE_TOKEN_HERE with your actual workspace token

  2. Save the file and restart Claude Desktop

Claude Code (CLI)

Use the Claude CLI to add the MCP server (don’t edit ~/.claude.json manually). This writes the config for you.

  1. Add the Linkbreakers MCP server:
Bash
claude mcp add --transport http linkbreakers https://mcp.linkbreakers.com
  1. Follow the Claude CLI prompts to set the Authorization header with your workspace token:
JavaScript
Authorization: Bearer YOUR_WORKSPACE_TOKEN_HERE
  1. Restart Claude Code

Continue (VSCode)

  1. Open VSCode and install the Continue extension

  2. Locate your configuration file at ~/.continue/config.json

  3. Add the MCP server configuration:

JSON
{
  "mcpServers": [
    {
      "name": "linkbreakers",
      "url": "https://mcp.linkbreakers.com",
      "headers": {
        "Authorization": "Bearer YOUR_WORKSPACE_TOKEN_HERE"
      }
    }
  ]
}
  1. Replace YOUR_WORKSPACE_TOKEN_HERE with your workspace token

  2. Restart VSCode

Other MCP clients

For other MCP-compatible clients, add these connection details:

  • Server URL: https://mcp.linkbreakers.com
  • Authentication: Bearer token in Authorization header
  • Protocol: MCP over HTTP/JSON

Refer to your client's documentation for specific configuration steps.

Step 3: Test the connection

After configuration, verify the MCP server is working:

In Claude Desktop

  1. Start a new conversation

  2. Type / to see available tools

  3. Look for Linkbreakers tools like:

    • links_create
    • links_list
    • metrics_workspace
    • events_list
  4. Test with a simple request: "List my links from Linkbreakers"

Claude should use the links_list tool and return your workspace links.

In Continue

  1. Open the Continue sidebar in VSCode
  2. Check the MCP tools list
  3. Test with a prompt: "Show me my Linkbreakers workspace analytics"

Troubleshooting connection issues

If tools don't appear:

  1. Verify token is correct - Check for typos or extra spaces
  2. Confirm token hasn't been deleted - Check API Tokens in Linkbreakers settings
  3. Ensure correct format - Authorization header must include "Bearer " prefix
  4. Restart your client completely - Some clients need a full restart
  5. Check logs - Your MCP client may show connection errors in logs

Available MCP tools

The Linkbreakers MCP server exposes these curated operations:

  • links_create - Create new shortened links
  • links_list - List all links with filtering options
  • links_get - Get details of a specific link
  • links_update - Update link metadata and settings
  • links_delete - Delete a link

Workflow steps

  • workflow_steps_list - List workflow steps for a link
  • workflow_steps_get - Get workflow step details
  • workflow_steps_delete - Delete a workflow step
  • workflow_steps_create_relationship - Connect workflow steps

Analytics and metrics

  • events_list - Retrieve event timeline and activity
  • metrics_workspace - Get workspace-level statistics

QR code designs

  • qrcodedesigns_create - Create a QR code design

QR code templates

  • qrtemplates_list - List QR code templates
  • qrtemplates_get - Get QR code template details
  • qrtemplates_create - Create a QR code template
  • qrtemplates_delete - Delete a QR code template

Media

  • media_create - Upload media assets
  • media_list - List media assets
  • media_get - Get media details
  • media_delete - Delete a media asset

Visitors

  • visitors_list - List visitors with filtering
  • visitors_get - Get visitor details

Custom domains

  • domains_create - Add custom domains
  • domains_list - List custom domains
  • domains_get - Get domain details
  • domains_check - Verify DNS and TLS status

Tags

  • tags_list - Search and list tags

Example use cases

Prompt:

"Create a shortened link for https://example.com/products with the slug 'spring-sale' and add the tag 'marketing'"

What happens:

  1. AI calls links_create tool with parameters
  2. MCP server proxies request to Linkbreakers API
  3. Link is created and short URL is returned
  4. AI formats the response for you

Analyze campaign performance

Prompt:

"Show me the click metrics for all links tagged 'marketing' in the last 30 days"

What happens:

  1. AI calls links_list with filters
  2. AI calls events_list for click data
  3. AI aggregates and analyzes the data
  4. AI presents insights and trends

Automate workflow creation

Prompt:

"Create a workflow for my newsletter link with a password gate, then redirect to a form, then to the content URL"

What happens:

  1. AI calls links_create for base link
  2. AI calls workflowsteps_create for each step
  3. AI calls workflowsteps_createrelationship to connect them
  4. Fully configured workflow is ready to use

Best practices

Token management

  • Create separate tokens for each device (e.g., "Claude Desktop - Work MacBook", "Continue - Home PC")
  • Use descriptive names to identify tokens later
  • Rotate tokens every 90 days for security
  • Delete unused tokens immediately when changing devices or clients

AI prompts

  • Be specific - "Create a link for X with slug Y" works better than "Make a link"
  • Include context - "Show analytics for the marketing campaign" is clearer than "Show stats"
  • Use filters - "List links created this week" vs "List links" returns focused results
  • Ask follow-ups - AI can chain multiple operations based on responses

Rate limiting

  • The MCP server inherits standard Linkbreakers API rate limits
  • Workspace tokens typically have higher limits than member tokens
  • If you hit rate limits, contact support for increased quotas
  • Consider batching requests in off-peak hours for large operations

Error handling

  • If a tool fails, AI will explain the error and may suggest fixes
  • Common errors include:
    • Invalid parameters - Check parameter format and required fields
    • Authentication failed - Token may be expired or revoked
    • Resource not found - ID may be incorrect or deleted
    • Rate limit exceeded - Wait before retrying

Architecture

The MCP server works as a stateless protocol translator:

Loading diagram...

Key characteristics:

  • Stateless - No database or caching required
  • Secure - All authentication handled by main API
  • Simple - Direct token pass-through, no validation needed
  • Fast - Minimal overhead, direct API proxy

Frequently asked questions

Is the MCP server free to use?

Yes, the MCP server at https://mcp.linkbreakers.com is provided free to all Linkbreakers users. Standard API rate limits apply based on your workspace plan.

Can I use member tokens instead of workspace tokens?

No, the MCP server only accepts workspace tokens. Member tokens are designed for personal API access and don't work with MCP.

What happens if my token is compromised?

If you suspect your token has been compromised:

  1. Go to Settings → API Tokens in Linkbreakers
  2. Delete the compromised token immediately
  3. Create a new token with a different name
  4. Update your MCP client configuration with the new token

The old token will stop working immediately after deletion.

Can multiple team members use the same workspace token?

While technically possible, we strongly recommend creating separate tokens for each team member. This provides:

  • Better audit trails (know who made which changes)
  • Easier token revocation (remove access for one person)
  • Reduced risk (compromised token affects only one user)

Does the MCP server store my data?

No. The MCP server is stateless and doesn't store any data. It:

  • Receives your request
  • Forwards it to Linkbreakers API with your token
  • Returns the API response
  • Keeps no logs or records of the data

All data storage and processing happens in the main Linkbreakers API.

What MCP clients are supported?

Any MCP-compatible client works, including:

  • Claude Desktop (official Anthropic client)
  • Continue (VSCode extension)
  • Custom MCP clients following the MCP specification

Can I use this with the Linkbreakers mobile app?

The MCP server is designed for desktop AI assistants, not mobile apps. Use the standard Linkbreakers API or mobile app for mobile access.

How do I report bugs or request features?

For MCP server issues:

For feature requests:

  • Contact support with your use case
  • Describe which API endpoints you need exposed

About the Author

LS

Laurent Schaffner

Founder & Engineer at Linkbreakers

Passionate about building tools that help businesses track and optimize their digital marketing efforts. Laurent founded Linkbreakers to make QR code analytics accessible and actionable for companies of all sizes.