Overview
Yardi Virtuoso MCP is an implementation of the Model Context Protocol (MCP) that enables AI agents to securely interact with Yardi systems, data sources, and operational tools. This guide provides a comprehensive explanation of the MCP protocol, describes how authentication operates with the Yardi Virtuoso MCP server, details the available MCP tools, and presents example agent interactions.
What Is the Model Context Protocol (MCP)?
The Model Context Protocol (MCP) is an open, standardized protocol introduced by Anthropic to connect AI assistants to external systems such as business tools, data repositories, and development environments. MCP offers a universal and secure method for AI models to discover and invoke tools and retrieve contextual data, eliminating the need for custom integrations for each system.
According to Anthropic, MCP operates as a client–server protocol. MCP servers are responsible for exposing tools, data, and resources, while MCP clients—such as AI agents or applications—discover and invoke these capabilities. Communication between clients and servers occurs over a structured, two-way interface, as opposed to ad-hoc APIs.
MCP is designed to replace fragmented integrations with a single open standard, reducing complexity and improving reliability when connecting AI systems to real-world data and actions.
Why MCP Matters for Yardi Virtuoso
Standardized AI access to Yardi data and workflows
Secure, auditable tool invocation by AI agents
Reduced integration overhead for new tools and services
Future-proof compatibility with MCP-aware AI platforms
Authentication with Yardi Virtuoso MCP Server and YardiOne
This section outlines the authentication process required for clients before they can invoke tools or access resources on the Yardi Virtuoso MCP server.
Authentication Overview
Yardi Virtuoso MCP uses YardiOne to authenticate between the MCP clients such as Claude, ChatGPT and Copilot and the MCP server to ensure that only authorized users can connect. Tool access is limited to permitted capabilities, and all interactions are traceable and auditable.
Authentication Flow (High Level)
Claude or ChatGPT app initiates a connection to the Yardi Virtuoso MCP server.
Claude presents the YardiOne screen for the user to enter authentication credentials.
Server validates credentials and applies authorization policies.
MCP session is established with scoped permissions.
MCP Tools Provided by Yardi Virtuoso
The Yardi Virtuoso MCP server exposes a collection of MCP tools that AI agents can dynamically discover and invoke. Each tool is described using MCP’s standardized schema, including its purpose, required inputs, and expected outputs.
Yardi Virtuoso MCP Tools
Here are the tools currently available within Yardi Virtuoso for Claude integration:
Work Order Management Tools
Create Work Order (rfm_connectai_create_work_order): Creates new work orders in your Yardi system. You can specify comprehensive details including brief and full descriptions, property information, unit assignment, priority levels, categories, and custom fields. This tool supports attachments and allows you to define access notes and completion status upfront.
Search Work Orders (rfm_connectai_search_work_orders): Performs advanced filtering and search across your work order database with extensive query capabilities. You can filter by due dates, creation dates, priority levels, statuses, categories, subcategories, assigned employees, vendors, properties, buildings, units, and tenants. The tool also supports searching by attachment presence, purchase orders, technician notes, and work order origin codes, with options to return up to 5000 records.
Mark Work Order as Complete (rfm_workorder_mark_work_order_as_complete): Closes completed work orders in the system by recording the completion date and adding final technician notes and descriptions. This ensures your maintenance tracking accurately reflects completed tasks.
Autocomplete and Entity Lookup Tools
Get Autocomplete Suggestions (vn_mcpframework_get_autocomplete): Provides intelligent autocomplete suggestions for various entity types including properties, vendors, tenants, units, banks, PM users, employees, accounts, buildings, floors, categories, prospects, and roles. You can search with free-text terms and specify whether you want exact or partial matches, with customizable result limits.
List Autocomplete Types (vn_mcpframework_list_autocomplete_types): Returns the available autocomplete entity types that can be used with the autocomplete tool, helping you understand what data entities can be looked up.
Yardi Virtuoso MCP Tools: User Prompts & Sample Output
Search Work Orders
User prompt: “Show me all open high-priority work orders for HVAC problems at Gardenview Terrace created in the last 30 days.”
Agent Output
There are 5 open high-priority HVAC work orders:
WO #48392 (Unit 204): No A/C; unit runs but isn’t cooling—due Feb 10.
WO #48401 (Building 2): Annual HVAC inspection needed—due Feb 28.
WO #48389 (Unit 108): Thermostat display frozen, temp control issue—was due Feb 8 (overdue).
WO #48405 (Unit 312): Outdoor condenser making grinding noise—due Feb 12.
WO #48398 (Unit 410): Ductwork air leak causing energy loss—due Feb
Summary: Five urgent HVAC issues need attention at Gardenview Terrace. The oldest request for Unit 108 is now overdue.