# eMaint MCP server

Source: https://agentic-cmms.klementine.xyz/integrations/emaint

> Connect eMaint CMMS to Claude, ChatGPT and other AI agents for work orders, PMs, assets, failure codes, inventory and condition data.

eMaint is a configurable CMMS used by manufacturers, food and drink plants, pharmaceutical sites and facility teams that take reliability seriously. It manages work orders, preventive maintenance, assets, inventory and purchasing, and it can sit next to condition monitoring tools, so readings and alarms become part of the maintenance record. Many eMaint sites have shaped it to their own process over the years, with custom fields and screens. That flexibility makes cross-site answers harder to get.

Agentic CMMS gives your eMaint system an MCP server. Reliability engineers and maintenance planners can ask Claude, ChatGPT or another AI agent about failures, PMs and condition data in plain language. An approved agent can also raise and update work. You connect eMaint once on the Connect page; the credentials are checked against eMaint before they are saved.

## What your AI agent can do with eMaint

- **Turn a reading into action.** "Which motors have a vibration reading above our alert level this week but no open work order?" The agent compares the latest readings with your limits and the open work.
- **Report failures properly.** "Count the failure codes on the extruder line for the last six months and show the top three causes." The agent lists the work orders behind each count.
- **Plan the PM week.** "Which PMs are due next week on second shift, and how many labour hours do they need?"
- **Raise work with the right codes.** "Create a corrective work order on Mixer M-4 for a seal leak, failure code LEAK, priority 2."
- **Keep inventory honest.** "Which parts were issued to work orders this month but are now below their reorder point?"
- **Support an audit.** "List every calibration PM on the QC lab instruments from last quarter, with the completion date and who signed it off."
- **Close out with detail.** "Complete my work order on the air dryer, add two hours of labour and record the filter part I used."

## eMaint with your other systems

Reliability work depends on data from many places. Connect eMaint to the systems around it and the agent can bring them together in one answer:

- Compare the asset readings in [AWS IoT SiteWise](/integrations/aws-iot-sitewise) with eMaint PM intervals to find PMs that come too early or too late.
- Link calibration work orders in eMaint with the change records and deviations in [MasterControl](/integrations/mastercontrol) before a quality audit.
- Match spare part usage in eMaint with purchase prices in [Sage X3](/integrations/sage-x3) to see the true parts cost of each production line.
- Add eMaint failure counts to the reliability dashboards your team keeps in [Tableau](/integrations/tableau).

## For reliability-led teams

eMaint customers often run structured programmes: failure codes, criticality, condition-based PMs and regular reviews. An agent is most useful when it keeps that structure and shows the records behind every number. Each person approves their own agent, owners and admins limit each eMaint tool to the right roles, and every call is written to the tamper-evident activity log.

## Questions and answers

### Can an AI agent use eMaint condition data to plan work?

An approved agent can read meter and condition readings stored on eMaint assets and compare them with limits and recent work. It can then suggest or, with write access, create the work order.

### Our eMaint is heavily customised. Will Claude understand our fields?

The agent reads the records with their field names and values, including the ones your team set up. You can tell it what a field means once in your conversation and it uses that from then on.

### Can ChatGPT build a failure report from eMaint without exporting data?

Yes. Ask for failure codes by asset or area over a period and the agent counts and summarises them for you, with the work order numbers behind each figure.