This is a library to interface with La Marzocco's "new" Home machines (currently only the Micra).

It's in experimentals stages and meant mainly to connect to the Micra, as for the other IoT enabled machines you can use the lmdirect library.

Standalone scripts are added to toggle coffee boiler and steam boiler respectively.

The scripts can be executed in any network-enabled machine, like a Synology NAS. This version is Docker enabled to ensure a up-to-date Python runtime.

To run the scripts in Synology NAS (general steps below):

1. Install Docker from Synology package center
2. Sync/copy this repo to a directory on the NAS you have access to, e.g. your home directory
3. cd into the copied repo on the NAS (requires SSH access) and build this into a Docker image (docker build -t lmcloud .)
4. Run the image in a new container either from the UI or with docker run

The scripts can be called with a simple docker exec command, for example:

docker exec lmcloud python3 toggle_coffee_boiler.py

Assuming the container is named lmcloud.

• lmlocalapi calls the new local API the Micra exposes, using the Bearer token from the customer cloud endpoint. However, this API currently only supports getting the config, and some status objects (like shottimer) over websockets, but does not support setting anything (to my knowledge). Local settings appear to only happen through Bluetooth connections. If La Marzocco updates the firmware or more endpoints are found this library will be updated to reflect those additional endpoints.
• lmcloud interacts with gw.lamarzocco.com to send commands. lmcloud can be initialized to only issue remote commands, or to initialize an instance of lmlocalapi for getting the current machine settings. This helps to avoid flooding the cloud API and is faster overall.
• lmbluetooth provides a bluetooth client to send settings to the machine via bluetooth

Because of that reason the config object self._config in the lmcloud instance without utilizing the local API will always at least be 10 seconds old. This is to avoid automatic property checks (e.g. from HomeAssistant) to spam the cloud API. If you really require a most recent config you can call the method get_config().

To run lmcloud you will first need to create a dict, containing clientId, clientSecret, username and password.

username and password are easy to get because those are the credentials you're using to sign into the La Marzocco Home app.

clientId and clientSecret are harder to get. The "easiest" way is to configure mitmproxy on a PC, configure your WiFi settings on your phone to use that proxy. Then log out of your La Marzocco App, sign in again and watch for a call to https://cms.lamarzocco.io/oauth/v2/token. The credentials you're looking for will be in that call. More of the calls are documented in remote_rest.

You then need to create a file called config.json looking like this:

{
    "client_id": "long_string",
    "client_secret": "long_string",
    "username": "email",
    "password": "password",
    "host": "192.168.1.123" # optional, only if you want the local API
}

Which then needs to be passed into a dictionary:

with open("config.json") as f:
    data = json.load(f)

creds = {
    "client_id": data["client_id"],
    "client_secret": data["client_secret"],
    "username": data["username"],
    "password": data["password"]
}

lmcloud exposes two classmethods: One to build an instance with the local API, and one to build an instance relying entirely on cloud functions.

It is initialized like this

lm_cloud = await LMCloud.create(creds) # cloud only
lm_cloud = await LMCloud.create_with_local_api(creds, ip=ip) # with local API

If you just want to run the local API you need the IP of your machine, the Port it is listening on (8081 by default), the Bearer token (communicationKey) used for local communication. You can obtain that key by inspecting a call to https://cms.lamarzocco.io/api/customer, while connected to mitmproxy (process above), or making a new (authenticated) call to that endpoint. lmcloud will also connect to this endpoint so you could also used a breakpoint there to get this key, which is stored in a key called communicationKey.

Then you can init the class with

lm_local_api = LMLocalAPI(ip, bearer)

The local API initiates a websocket connection to

http://{IP}:8081/api/v1/streaming

The packets which are received on that WebSocket are documented in websockets

If WebSockets are enabled the shot timer becomes available to use, however as long as the library is running in WebSocket mode, the App will no longer be able to connect.

To use WebSockets start the integration with

lm = await LMCloud.create_with_local_api(creds, <IP>, port=8081, use_websockets=True)

while True:
    print(lm._lm_local_api.active_brew) # is a brew running at the moment
    print(lm._lm_local_api.active_brew_duration) # the current shot timer returned by the machine

Some commands, like turning the machine on and off are always sent through bluetooth whenever possible. The available bluetooth characteristics are described in bluetooth_characteristics. The class LMBluetooth discovers any bluetooth devices with Micra in the name and connects to it. Then we can send local bluetooth commands.

To use Bluetooth you can either init LMCloud with

    lm = await LMCloud.create_with_local_api(creds, <IP>, port=8081, use_bluetooth=True)

or even init the Bluetooth client standalony

    lm_bt = await LMBluetooth.create(username, serial_number, token)

The token is the same token you need to initialize the local API, which you need to get from LM's cloud once. The serial number is your machine's serial number and the username is the email of your LaMarzocco account.