Connect ChatGPT and Claude to your CMMS with MCP
The MaintenEase MCP server gives compatible AI clients a controlled way to read maintenance data and create work orders after a user signs in and approves access. Your database credentials are never shared with the model.
Production MCP endpoint
https://wwgljhpuulhljumrhscg.supabase.co/functions/v1/mcp. Streamable HTTP transport with OAuth 2.1 with PKCE.
Review work orders: list_work_orders
Read access. List the signed-in company's work orders and optionally filter them by status.
Create a work order: create_work_order
Write access. Create an authorized work order with a title, description, priority, due date, and optional asset.
Find tracked assets: list_assets
Read access. Browse equipment and assets available to the signed-in user's company.
Find buildings and rooms: list_locations
Read access. List the sites, buildings, rooms, and other locations the company tracks.
Review incoming requests: list_maintenance_requests
Read access. Read the maintenance request inbox before turning a valid request into tracked work.
Human-approved access
OAuth consent keeps the user in control of whether an AI client can connect to MaintenEase.
Signed-in identity
Every tool call carries the authorized user's identity instead of relying on a shared organization credential.
Tenant-scoped results
Database row-level security limits records to the signed-in user's company.
Narrow, declared tools
The client can call only the maintenance actions published by the MCP server and their validated inputs.
Add the MaintenEase MCP server
Use the endpoint or server card in an AI client that supports remote MCP servers and OAuth.
Sign in to MaintenEase
The client opens the MaintenEase authorization flow without receiving the user's password.
Review and approve access
The user confirms the connection before the client receives an access token.
Use maintenance tools
The AI client can select an available tool while MaintenEase continues to enforce user and company permissions.
MCP server questions
What is the MaintenEase MCP server?
It is a Model Context Protocol service that publishes controlled CMMS tools for compatible AI clients. It can list work orders, assets, locations, and maintenance requests and can create a work order after user authorization.
Can ChatGPT or Claude connect to MaintenEase?
They can connect when the relevant product, plan, and workspace policy support remote OAuth-authenticated MCP servers. Availability and setup controls can vary by AI client.
Does the AI client receive database credentials?
No. The client receives a revocable OAuth access token after user consent. MaintenEase applies the signed-in user's permissions and tenant-scoped data rules to every tool call.
Can an AI client create work orders?
Yes. The create_work_order tool accepts a narrow, validated set of fields. The work order is created for the signed-in user's company and records that user as its creator.
Where can an AI crawler find the machine-readable details?
Use the MCP server card, the clean Markdown version of this page, llms.txt, llms-full.txt, and the structured AI catalog at /api/ai.json.