Gaggiuino MCP Server

What is Gaggiuino MCP Server about?

Gaggiuino MCP Server enables AI applications to retrieve the current state of a Gaggia Classic espresso machine running the open‑source Gaggiuino controller, obtain the most recent shot identifier, and fetch full telemetry data for any specific shot.

How to use Gaggiuino MCP Server?

Start the server (Python) with the command shown in the Claude Desktop configuration:

uv --directory /ABSOLUTE/PATH/TO/PARENT/FOLDER/gaggiuino-mcp run gaggiuino.py

Connect a client that speaks the Model Context Protocol and issue one of the defined toolset commands (getStatus, getLatestShotId, getShotData).
Install automatically for Claude Desktop via Smithery:

npx -y @smithery/cli install @AndrewKlement/gaggiuino-mcp --client claude

Use the JSON snippets in Claude Desktop settings to launch the server.

Key features of Gaggiuino MCP Server

Real‑time retrieval of machine status (temperature, pressure, sensor states).
Immediate access to the latest espresso shot identifier.
Detailed shot telemetry download (flow rate, pressure, temperature over time).
Simple HTTP‑based toolset commands usable by any MCP‑compatible AI client.
Designed for local‑network deployment with minimal configuration.

Use cases of Gaggiuino MCP Server

AI‑driven espresso analysis – feed shot data into language models for automatic quality assessment.
Dashboard visualisation – display live machine metrics and historic shot curves on a web UI.
Remote monitoring – allow barista assistants to check machine status from a phone.
Research & development – collect large volumes of shot telemetry for profile optimisation.
Automation – trigger alerts or adjustments when temperature or pressure drifts.

FAQ from the Gaggiuino MCP Server

Q: Do I need any special hardware? A: Only a Gaggia Classic equipped with the open‑source Gaggiuino controller.

Q: Which programming languages can talk to the server? A: Any language that can send MCP JSON requests (Python, JavaScript, etc.).

Q: Is the server secure for internet exposure? A: It is intended for local‑network use; expose it publicly only behind proper authentication and TLS.

Q: How are shot IDs generated? A: The Gaggiuino firmware increments an integer each time a shot completes.

Q: Can I customize the telemetry resolution? A: Yes, adjust the sampling interval in the Gaggiuino firmware configuration.

gaggiuino-mcp

Gaggiuino MCP Server This is a lightweight Model Context Protocol (MCP) server built for Gaggiuino, the open-source espresso machine controller for the Gaggia Classic. It is designed to integrate easily AI clients that want to display or analyze data from the Gaggiuino system in real time.

This is an article I wrote about this project https://medium.com/@a.klement.w/dialing-in-espresso-using-mcp-server-0207819fe93a

Allows connected clients to:

Retrieve the current machine status
Access the latest shot ID
Fetch shot data for a specified id

Features

📊 Real-time access to shot telemetry

🌐 Designed for local network access

How to use

Using Claude Desktop

{
  "mcpServers": {
      "gaggiuino": {
          "command": "uv",
          "args": [
              "--directory",
              "/ABSOLUTE/PATH/TO/PARENT/FOLDER/gaggiuino-mcp",
              "run",
              "gaggiuino.py"
          ]
      }
  }
}

Installing via Smithery

To install Gaggiuino Espresso Machine Server for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @AndrewKlement/gaggiuino-mcp --client claude

Available Toolsets

Toolset	Description
`getLatestShotId`	Get latest espresso shot id
`getShotData`	Get espresso shot data for an id. Args: id: Shot id
`getStatus`	Get espresso machine status

Example

User:

analyze the latest espresso shot data

Claude:

I'll retrieve and analyze the latest espresso shot data for you. Let me first check the status of your espresso machine and then get the latest shot information. getStatus Now I'll get the ID of the latest espresso shot: getLatestShotId Let me retrieve the data for shot #188: getShotData Based on the data retrieved, I'll now analyze your latest espresso shot (ID #188). Here's my analysis:

Machine Status and Shot Overview Your machine is currently idle at 57.2°C with -0.1 bar pressure. The shot analyzed was taken on April 19, 2025 (based on the timestamp), with a total duration of 26.5 seconds (265 datapoints at 0.1s intervals).

Profile Information You used the "Leva 9 LR v0.5" profile, which is a multi-phase profile with declining pressure targeting 40g of output. The profile has

A flow-control preinfusion at 7 ml/s
Step-down to 3.5 ml/s flow
Pre-extraction at 1.5 ml/s flow with 3 bar pressure restriction
Pressure ramp to 9 bar over 2 seconds
A series of gradually declining pressure steps (8.8→6.0 bar)

Shot Analysis Preinfusion and Extraction Phases

Preinfusion (0-5s): Started with a high flow rate (6-7 ml/s) with minimal pressure (0.3-0.6 bar), allowing the puck to saturate.
First Pressure Rise (5-10s): Pressure gradually built to ~2.5 bar while flow decreased to ~1.5 ml/s.
Pressure Profile Phase (10-15s): Pressure held around 3 bar with flow reduced to ~0.8 ml/s.
Pressure Ramp (15-20s): Pressure increased from 3 to 8.9 bar over ~5 seconds following the programmed profile.
Declining Pressure Phase (20-26.5s): Pressure gradually declined from ~8.5 to 6.4 bar following the profile steps.

Key Measurements

Final Yield: 38g (based on final shot weight)
Brew Ratio: Unable to determine (coffee dose not provided)
Temperature Stability: Excellent - maintained at 92.8-94.1°C throughout extraction
First Drops: Appeared around 11-12 seconds (~75% into preinfusion)
Peak Flow Rate: 7.2 ml/s (at the beginning)
Peak Pressure: 8.9 bar (at ~20 seconds)

Extraction Dynamics

Flow Rate: Started high (7 ml/s), stepped down to 3.5 ml/s then 1.5 ml/s during preinfusion as programmed, then stabilized at ~2.8 ml/s during extraction.
Water Delivery: Total water pumped was 87.7 ml, with 38g in the cup, suggesting significant water retention in the puck.
Pressure Profile: The pressure curve shows excellent progression from preinfusion through extraction, with a smooth transition to declining pressure.

Performance Assessment This appears to be a technically well-executed shot with good adherence to the programmed profile:

Temperature Control: Excellent stability around 93-94°C
Pressure Curve: Follows the programmed profile with smooth transitions
Extraction Flow: Consistent during main extraction phase
Weight Output: Achieved 38g (close to the 40g target)

Would you like me to provide more specific recommendations about any aspect of this shot, or is there a particular parameter you'd like me to analyze in more detail?