How to Ask AI About Your Restoration Projects (magicplan MCP)

Restoration

5 min read

Your crew already captured the job: room measurements, photos, moisture readings, affected areas, forms, and notes. Yet before an adjuster or homeowner call, someone still has to open the project and search through everything.

Model Context Protocol (MCP) offers a faster way to get answers from that field record. It gives an AI assistant controlled access to the project data stored in magicplan, so you can ask questions such as:

Is the affected room dry yet?
What happened on this water loss?
Which rooms are missing documentation?
What should I know before calling the homeowner?

Instead of relying on general AI knowledge, the assistant uses the information your team captured on site.

What Is Model Context Protocol?

Model Context Protocol is an open standard that connects AI applications to external tools and data sources.

Without an MCP connection, an AI assistant cannot see your open jobs, floor plans, photos, moisture readings, or estimates. With magicplan MCP, supported assistants such as ChatGPT and Claude can read workspace data and reason over it conversationally.

This is different from a traditional API:

MCP is designed for asking AI questions about your project data.
The REST API is designed for developers building integrations and automations.

To explore the technical options, see the magicplan MCP documentation or the magicplan REST API documentation.

Why MCP Is Useful for Restoration Teams

Restoration data only becomes useful when its context stays intact. A moisture reading needs a room. A photo needs an affected surface. A drying update needs previous readings for comparison.

The crew on site understands the loss, but the office often has to reconstruct that knowledge later from scattered files and incomplete notes. MCP helps project managers, estimators, and owners retrieve what the field already documented without manually searching the entire project.

That can mean fewer clarification calls, quicker job reviews, and stronger handoffs from mitigation to estimating and reconstruction.

What Can You Ask AI About a magicplan Project?

A project manager preparing for an adjuster call could ask:

“Give me a quick summary of the Reed water loss.”

An estimator could ask:

“Which affected rooms have enough documentation to support the scope?”

A technician could ask:

“Summarize the drying progress in the basement.”

An operations manager could ask:

“Which projects have not moved recently?”

You can also ask the assistant to identify missing documentation, review photos, summarize estimates, or turn drying data into a report. For water-loss workflows, magicplan Moisture Mapping keeps readings connected to their locations and tracks them across visits.

The quality of every answer still depends on the quality of the field record. MCP makes captured information easier to use; it cannot recover details that were never documented.

How to Connect magicplan to an AI Assistant

You need access to a compatible magicplan workspace and an AI client that supports MCP.

1. Find your workspace credentials

In the magicplan Cloud, open the workspace settings and go to API & Integrations. Locate the workspace Customer ID and API key.

2. Add the magicplan MCP connector

Add this server address through your AI assistant’s connector settings:

https://cloud.magicplan.app/mcp

3. Authorize the connection

Enter the Customer ID and API key in the authorization screen.

Never paste an API key into a normal AI conversation. If you believe a credential has been exposed, rotate it.

4. Ask about a project

Once connected, refer to a project by name and ask your question in plain language. The assistant selects the available tools it needs to retrieve the relevant information.

For detailed instructions, follow the official magicplan MCP setup guide.

What Data Can magicplan MCP Read?

Depending on what your team has captured, the assistant can work with:

Projects, floor plans, measured areas, and forms
Photos and moisture history linked to the property
Attached files
Estimates, line items, and totals

Unlike uploading one static PDF to an AI chat, MCP lets the assistant reason across different parts of the same job record. The existing magicplan reporting workflow remains useful when you need standard outputs that combine floor plans, photos, readings, and equipment documentation.

Is magicplan MCP Read-Only?

Yes. The current magicplan MCP server is read-only. An AI assistant can inspect and analyze project information, but it cannot edit, move, or delete the underlying data.

The field record remains the source of truth. AI helps your team understand and package it without introducing unexpected changes to the project.

Use Skills for Repeatable AI Workflows

You can ask questions with ordinary prompts, but recurring tasks benefit from consistent instructions. A project summary, photo review, drying report, and adjuster handoff each require a different process and output.

The magicplan AI Skills Library provides ready-made workflows for tasks including:

Project summaries
Drying reports
Photo reviews
Mitigation scopes
Adjuster handoffs
Workspace and estimate-status reviews

Skills reduce the need for every employee to invent the right prompt. They help teams repeat the same process across jobs while using the information already stored in magicplan.

MCP, Skills, or REST API: Which Should You Use?

Use MCP when you want an AI assistant to answer questions about workspace data.

Use Skills when you want the assistant to follow a defined restoration workflow and produce a consistent result.

Use the REST API when developers need to connect systems, manage data programmatically, or build automations. You can also review magicplan’s existing options in the integrations directory.

Frequently Asked Questions

What does MCP mean in simple terms?

MCP stands for Model Context Protocol. It lets an AI assistant connect to external data and tools. With magicplan, it allows a supported assistant to read authorized workspace information when answering project questions.

Can ChatGPT or Claude read my magicplan projects?

Yes, after a supported MCP-capable assistant has been connected and authorized for the relevant workspace.

Can AI change my magicplan projects through MCP?

No. The current connection is read-only and cannot modify project data.

Do I need to write API calls?

No. MCP is designed for conversational use. You ask questions in plain language. Developers who need programmatic access should use the REST API.

Where can I find ready-made restoration workflows?

The magicplan AI Skills Library includes workflows for project summaries, drying reports, photo reviews, scopes, handoffs, and workspace-level reporting.

Turn Field Data Into Answers

The technician should not have to explain the same loss repeatedly. The project manager should not have to inspect every photo before a call. The owner should not have to open every job to see where work has stalled.

Your crew already captured the hard part: what happened on site. MCP gives the rest of the business a faster way to use that record for answers, summaries, reports, and decisions.

Start with the magicplan MCP documentation, then explore the magicplan AI Skills Library for repeatable restoration workflows.

The goal is not to make AI the expert on the job. The crew is the expert. MCP simply makes their knowledge easier to use.