MCP Server

Let Your AI Assistant See Your Data

Connect Claude Desktop, Cursor, and other MCP-compatible tools to Administrate. Ask questions about clients, workflows, and errors without leaving your AI interface.

Claude DesktopCursorSame security as API
MCP Server Configuration

Why Your AI Assistant Needs Data Access

AI is more useful when it can see what you're working with.

AI Assistants Can't See Your Data

You're asking Claude about a client's workflow failures, but it can't access Administrate. You end up copy-pasting data into the chat, losing context and wasting time.

Context Switching Kills Flow

Jumping between your AI assistant and Administrate to look up client info, check error counts, or find workflow names. Every switch breaks your train of thought.

Manual Data for AI Analysis

Your AI assistant could help analyze trends and spot issues, but first you have to export data, format it, and paste it in. The friction kills the value.

AI-Powered Data Access

Give your AI assistant direct access to monitoring data.

MCP Protocol Support

Works with Claude Desktop, Cursor, and other MCP-compatible AI tools. Your assistant can query Administrate data directly.

Query Clients and Instances

Ask your AI to 'list all clients with failures this week' or 'show me the workflows for Client X'. It fetches the data directly from Administrate.

Access Metrics and Errors

Let your AI analyze execution trends, identify problem workflows, or summarize LLM costs. All without leaving your AI interface.

Same Permissions as API

MCP uses the same API tokens with the same permission levels. Read-only tokens stay read-only. IP allowlists are respected.

Real-Time Data Access

Your AI assistant gets current data, not stale exports. Ask about today's errors and get today's answers.

Connect Multiple Accounts

If you manage multiple Administrate accounts (for different agencies or clients), connect them all. Your AI can query any of them.

Simple Setup, Powerful Access

Connect your AI tools and start asking questions.

MCP Configuration

MCP Configuration

Simple setup with your API token and server URL

AI Assistant Integration

AI Assistant Integration

Query Administrate data directly from Claude or Cursor

Data Access

Data Access

Clients, instances, workflows, and errors available to your AI

Connect Your AI in Minutes

1

Create API Token

Generate a token with read or write access

2

Configure MCP Client

Add the server URL and token to Claude Desktop or Cursor

3

Start Asking Questions

Your AI can now query clients, workflows, and errors

Frequently Asked Questions

What is MCP?
MCP (Model Context Protocol) is a standard for connecting AI assistants to external data sources. It lets tools like Claude Desktop and Cursor access your Administrate data directly, so you can ask questions about clients, workflows, and errors without leaving your AI interface.
Which AI tools support MCP?
Claude Desktop, Cursor, and other MCP-compatible AI tools. The protocol is gaining adoption, so more tools are being added. Check if your preferred AI tool supports MCP servers.
Is this secure?
Yes. MCP uses the same authentication as the REST API - bearer tokens with permission levels. You can create a read-only token for AI access, set IP restrictions, and revoke tokens at any time. Your AI assistant can only see what the token allows.
What data can my AI access?
Everything available through the API: clients, n8n instances, workflows, executions, errors, LLM providers, and costs. The AI can query, filter, and analyze this data based on your questions.
Do I need a paid plan?
MCP server access is available on Pro and Enterprise plans. The Free plan includes API access but not MCP.
How do I set it up?
Create an API token in Administrate, then add the MCP server configuration to your AI tool. For Claude Desktop, this goes in your claude_desktop_config.json. Documentation with exact steps is available in the Developers section.

Make Your AI Assistant Smarter

Connect to Administrate and let your AI help you manage clients and workflows.