mcpchatting

A short, two days Hackathon #ifgiHACK25 project. (Presentation Video)

mcpchatting is based on the following projects:

• QGIS MCP
• MCP Chatbot

The mcpchatting system has been enhanced with expanded language model support and a built-in chat interface. This integrated solution removes dependency on external chat platforms, providing users with unrestricted access(using their own API keys) without the constraints typically associated with free-plan services like Claude. Moreover, it now uses Retrieval-Augmented Generation. It combines large language models (LLMs) with external knowledge sources (in this case QGIS docs) so that the LLM has more knowledge about e.g. QGIS functions. So far it was only tested with ChatGPT.

Prerequisites

• Python 3.10 or newer
• pip (Python package installer)
• Node.js and npm (for NPX commands)
• uv (Python package installer and environment manager) for the QGIS server

Installation

1. Start Generation Here

If you're on Mac, please install uv as

brew install uv

On Windows Powershell:

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Otherwise installation instructions are on their website:Install uv

2. Qgis MCP Plugin

You need to copy the folder qgis_mcp_plugin and its content on your QGIS profile plugins folder.

You can get your profile folder in QGIS going to menu Settings -> User profiles -> Open active profile folder Then, go to Python/plugins and paste the folder qgis_mcp_plugin.

On a Windows machine the plugins folder is usually located at:C:\Users\USER\AppData\Roaming\QGIS\QGIS3\profiles\default\python\plugins

and on MacOS: ~/Library/Application\ Support/QGIS/QGIS3/profiles/default/python/plugins

Then close QGIS and open it again. Go to the menu option Plugins -> Installing and Managing Plugins, select the All tab and search for "QGIS MCP", then mark the QGIS MCP checkbox.

3. Clone the Repository

git clone https://github.com/whlan02/Qgis_mcpChatting
cd Qgis_mcpChatting

4. Set Up a Virtual Environment (Recommended)

python -m venv venv

# Activate the virtual environment

# On Windows:
venv\Scripts\activate
# On macOS/Linux:
source venv/bin/activate

5. Install Required Python Packages

pip install -r requirements.txt

6. Configure Servers

Update the servers_config_Example.json file to point to the correct directories for your system:

{
  "mcpServers": {
      "filesystem": {
          "command": "npx",
          "args": [
              "-y",
              "@modelcontextprotocol/server-filesystem",
              "/ABSOLUTE/PATH/TO/YOUR/WORKING/DIRECTORY" 
          ]
      },
      "qgis": {
          "command": "uv",
          "args": [
              "--directory",
              "/ABSOLUTE/PATH/TO/PARENT/REPO/FOLDER/qgis_mcp",
              "run",
              "qgis_mcp_server.py"
          ]
      }
  }
}

After you have configured the servers_config_Example.json file, you should rename it to servers_config.json. Otherwise, the QGIS MCP Server will not start.

Usage

1. Start the QGIS MCP Server

1. In QGIS, go to plugins -> QGIS MCP-> QGIS MCP
2. Click "Start Server"

2. Start the QGIS MCP Client

In the root folder of the project, run:

python main.py

3. Choose LLM Model and set API Key

Only OpenAI and Deepseek are tested so far.