TIDY's MCP server lets any MCP-compatible AI agent manage your properties. Claude Desktop, Claude Code, Cursor, Codex CLI, VS Code with Copilot — if it supports MCP, it can now schedule cleanings, report maintenance, and coordinate turnovers through TIDY.
Setup takes under 5 minutes. Here's how.
What is MCP?
Model Context Protocol (MCP) is the open standard for connecting AI agents to external tools and services. Instead of hard-coding API calls, your agent discovers available tools through MCP's protocol and calls them as needed. This means your agent doesn't need to know TIDY's API schema in advance — it learns what's available automatically.
Step 1: Get Your API Token
You need a TIDY account and API token. If you don't have one:
- Create a free TIDY account
- Get your token via the login endpoint:
curl -X POST https://api.tidy.com/v1/customer/login \ -H "Content-Type: application/json" \ -d '{"email": "you@example.com", "password": "your-password"}' - Save the returned token for the next step.
Step 2: Add the MCP Server
Choose your platform:
Claude Code
claude mcp add tidy -- npx @tidyapp/mcp-serverSet your token as an environment variable: export TIDY_API_TOKEN=your-token
Claude Desktop
Go to Settings > Developer > Edit Config and add:
{
"mcpServers": {
"tidy": {
"command": "npx",
"args": ["@tidyapp/mcp-server"],
"env": {
"TIDY_API_TOKEN": "your-token"
}
}
}
}Cursor
Go to Settings > MCP and add the same JSON configuration as Claude Desktop.
VS Code (Copilot)
Create .vscode/mcp.json in your project:
{
"servers": {
"tidy": {
"command": "npx",
"args": ["@tidyapp/mcp-server"],
"env": {
"TIDY_API_TOKEN": "your-token"
}
}
}
}Codex CLI
codex mcp add tidy -- npx @tidyapp/mcp-serverStep 3: Test It
Open your agent and try:
"Use TIDY to list my addresses"
If you don't have any addresses yet:
"Use TIDY to add a new property at 123 Main St, Miami FL 33101. Gate code is #1234, park in the driveway."
Then try something more practical:
"Schedule a 2-hour cleaning next Tuesday afternoon at 123 Main St"
What Your Agent Can Do
Once connected, your agent has access to 5 MCP tools:
- message_tidy — The primary interface. Send any natural language request.
- get_status — Check status of a request or job.
- list_addresses — See all your properties.
- list_bookings — View upcoming and past cleanings.
- list_issues — Track maintenance issues.
The message_tidy tool is the most powerful — it accepts natural language and TIDY figures out what to do. Your agent can say "the bathroom faucet is leaking" and TIDY will create a maintenance issue, categorize it, and optionally assign it for resolution.
Bonus: Add Agent Skills Too
MCP gives your agent the tools. Skills give it the knowledge. For the best experience, install a TIDY agent skill alongside the MCP server:
npx skills add TIDYAPP/tidy-agent-skillsThis teaches your agent TIDY-specific workflows, like how to handle a guest turnover or set up recurring maintenance.
Learn More
- MCP Server landing page — Platform-specific setup details
- Full MCP documentation — Tool schemas, troubleshooting, advanced config
- Developer Hub — All 5 ways to connect to TIDY