Skip to main content

Add-on configuration

Each add-on is stored in a folder. The file structure looks like this:


Translation files, config and build all support .json, .yml and .yaml as the file type.

To keep it simple all examples use .yaml

Add-on script

As with every Docker container, you will need a script to run when the container is started. A user might run many add-ons, so it is encouraged to try to stick to Bash scripts if you're doing simple things.

All our images also have bashio installed. It contains a set of commonly used operations and can be used to be included in add-ons to reduce code duplication across add-ons, therefore making it easier to develop and maintain add-ons.

When developing your script:

  • /data is a volume for persistent storage.
  • /data/options.json contains the user configuration. You can use Bashio to parse this data.

TARGET="$(bashio::config 'target')"

So if your options contain

{ "target": "beer" }

then there will be a variable TARGET containing beer in the environment of your bash file afterwards.

Add-on Dockerfile

All add-ons are based on the latest Alpine Linux image. Home Assistant will automatically substitute the right base image based on the machine architecture. Add tzdata if you need to run in a different timezone. tzdata Is is already added to our base images.


# Install requirements for add-on
apk add --no-cache \

# Copy data for add-on
RUN chmod a+x /

CMD [ "/" ]

If you don't use local build on the device or our build script, make sure that the Dockerfile also has a set of labels that include:

io.hass.version="VERSION" \
io.hass.type="addon" \

It is possible to use your own base image with build.yaml or if you do not need support for automatic multi-arch building you can also use a simple docker FROM. You can also suffix the Dockerfile with the specific architecture to use a specific Dockerfile for a particular architecture, i.e. Dockerfile.amd64.

Build args

We support the following build arguments by default:

BUILD_FROMHolds the image for dynamic builds or buildings over our systems.
BUILD_VERSIONAdd-on version (read from config.yaml).
BUILD_ARCHHolds the current build arch inside.

Add-on configuration

The configuration for an add-on is stored in config.yaml.

name: "Hello world"
version: "1.1.0"
slug: folder
description: >-
"Long description"
- amd64
url: "website with more information about the add-on (e.g., a forum thread for support)"
123/tcp: 123
- config:rw
- ssl
image: repo/{arch}-my-custom-addon

Avoid using config.yaml as filename in your add-on for anything other than the add-on configuration. The Supervisor does a recursively search for config.yaml in the add-on repository.

Required configuration options

namestringThe name of the add-on.
versionstringVersion of the add-on. If you are using a docker image with the image option, this needs to match the tag of the image that will be used.
slugstringSlug of the add-on. This needs to be unique in the scope of the repository that the add-on is published in and URI friendly.
descriptionstringDescription of the add-on.
archlistA list of supported architectures: armhf, armv7, aarch64, amd64, i386.

Optional configuration options

machinelistDefault is support of all machine types. You can configure the add-on to only run on specific machines. You can use ! before a machine type to negate it.
urlurlHomepage of the add-on. Here you can explain the add-on and options.
startupstringapplicationinitialize will start the add-on on setup of Home Assistant. system is for things like databases and not dependent on other things. services will start before Home Assistant, while application is started afterwards. Finally once is for applications that don't run as a daemon.
webuistringA URL for the web interface of this add-on. Like http://[HOST]:[PORT:2839]/dashboard, the port needs the internal port, which will be replaced with the effective port. It is also possible to bind the protocol part to a configuration option with: [PROTO:option_name]://[HOST]:[PORT:2839]/dashboard and it's looked up if it is true and it's going to https.
bootstringautoauto start at boot is controlled by the system. manual configures the add-on to only be started manually.
portsdictNetwork ports to expose from the container. Format is "container-port/type": host-port. If the host port is null then the mapping is disabled.
ports_descriptiondictNetwork ports description mapping. Format is "container-port/type": "description of this port". Alternatively use Port description translations.
host_networkboolfalseIf true, the add-on runs on the host network.
host_ipcboolfalseAllow the IPC namespace to be shared with others.
host_dbusboolfalseMap the host D-Bus service into the add-on.
host_pidboolfalseAllow the container to run on the host PID namespace. Works only for not protected add-ons. Warning: Does not work with S6 Overlay. If need this to be true and you use the normal add-on base image you disable S6 by overriding /init. Or use an alternate base image.
host_utsboolfalseUse the hosts UTS namespace.
deviceslistDevice list to map into the add-on. Format is: <path_on_host>. E.g., /dev/ttyAMA0
homeassistantstringPin a minimum required Home Assistant Core version for the add-on. Value is a version string like 2022.10.5.
hassio_rolestrdefaultRole-based access to Supervisor API. Available: default, homeassistant, backup, manager or admin
hassio_apiboolfalseThis add-on can access the Supervisor's REST API. Use http://supervisor.
homeassistant_apiboolfalseThis add-on can access the Home Assistant REST API proxy. Use http://supervisor/core/api.
docker_apiboolfalseAllow read-only access to the Docker API for the add-on. Works only for not protected add-ons.
privilegedlistPrivilege for access to hardware/system. Available access: BPF, DAC_READ_SEARCH, IPC_LOCK, NET_ADMIN, NET_RAW, PERFMON, SYS_ADMIN, SYS_MODULE, SYS_NICE, SYS_PTRACE, SYS_RAWIO, SYS_RESOURCE or SYS_TIME.
full_accessboolfalseGive full access to hardware like the privileged mode in Docker. Works only for not protected add-ons. Consider using other add-on options instead of this, like devices. If you enable this option, don't add devices, uart, usb or gpio as this is not needed.
apparmorbool/stringfalseEnable or disable AppArmor support. If it is enabled, you can also use custom profiles with the name of the profile.
maplistList of Home Assistant directories to bind mount into your container. Possible values: homeassistant_config, addon_config, ssl, addons, backup, share, media, and all_addon_configs. Defaults to ro, which you can change by adding :rw to the end of the name.
environmentdictA dictionary of environment variables to run the add-on with.
audioboolfalseMark this add-on to use the internal audio system. We map a working PulseAudio setup into the container. If your application does not support PulseAudio, you may need to install: Alpine Linux alsa-plugins-pulse or Debian/Ubuntu libasound2-plugins.
videoboolfalseMark this add-on to use the internal video system. All available devices will be mapped into the add-on.
gpioboolfalseIf this is set to true, /sys/class/gpio will map into the add-on for access to the GPIO interface from the kernel. Some libraries also need /dev/mem and SYS_RAWIO for read/write access to this device. On systems with AppArmor enabled, you need to disable AppArmor or provide your own profile for the add-on, which is better for security.
usbboolfalseIf this is set to true, it would map the raw USB access /dev/bus/usb into the add-on with plug&play support.
uartboolfalseDefault false. Auto mapping all UART/serial devices from the host into the add-on.
udevboolfalseDefault false. Setting this to true gets the host udev database read-only mounted into the add-on.
devicetreeboolfalseIf this is set to true, /device-tree will map into the add-on.
kernel_modulesboolfalseMap host kernel modules and config into the add-on (readonly) and give you SYS_MODULE permission.
stdinboolfalseIf enabled, you can use the STDIN with Home Assistant API.
legacyboolfalseIf the Docker image has no labels, you can enable the legacy mode to use the config data.
optionsdictDefault options value of the add-on.
schemadictSchema for options value of the add-on. It can be false to disable schema validation and options.
imagestringFor use with Docker Hub and other container registries. This should be set to the name of the image only (E.g,{arch}-addon-example). If you use this option, set the active docker tag using the version option.
codenotarystringFor use with Codenotary CAS. This is the E-Mail address used to verify your image with Codenotary (E.g, [email protected]). This should match the E-Mail address used as the signer in the add-on's extended build options
timeoutinteger10Default 10 (seconds). The timeout to wait until the Docker daemon is done or will be killed.
tmpfsboolfalseIf this is set to true, the containers /tmp uses tmpfs, a memory file system.
discoverylistA list of services that this add-on provides for Home Assistant. Currently supported: mqtt, matter and otbr
serviceslistA list of services that will be provided or consumed with this add-on. Format is service:function and functions are: provide (this add-on can provide this service), want (this add-on can use this service) or need (this add-on needs this service to work correctly).
auth_apiboolfalseAllow access to Home Assistant user backend.
ingressboolfalseEnable the ingress feature for the add-on.
ingress_portinteger8099For add-ons that run on the host network, you can use 0 and read the port later via the API.
ingress_entrystring/Modify the URL entry point.
ingress_streamboolfalseWhen enabled, requests to the add-on are streamed
panel_iconstringmdi:puzzleMDI icon for the menu panel integration.
panel_titlestringDefaults to the add-on name, but can be modified with this option.
panel_adminbooltrueMake the menu entry only available to users in the admin group.
backupstringhothot or cold. If cold, the supervisor turns the add-on off before taking a backup (the pre/post options are ignored when cold is used).
backup_prestringCommand to execute in the context of the add-on before the backup is taken.
backup_poststringCommand to execute in the context of the add-on after the backup was taken.
backup_excludelistList of files/paths (with glob support) that are excluded from backups.
advancedboolfalseSet this to true to require the user to have enabled "Advanced" mode for it to show.
stagestringstableFlag add-on with follow attribute: stable, experimental or deprecated. Add-ons set to experimental or deprecated will not show up in the store unless the user enables advanced mode.
initbooltrueSet this to false to disable the Docker default system init. Use this if the image has its own init system (Like s6-overlay). Note: Starting in V3 of S6 setting this to false is required or the addon won't start, see here for more information.
watchdogstringA URL for monitoring the add-on health. Like http://[HOST]:[PORT:2839]/dashboard, the port needs the internal port, which will be replaced with the effective port. It is also possible to bind the protocol part to a configuration option with: [PROTO:option_name]://[HOST]:[PORT:2839]/dashboard and it's looked up if it is true and it's going to https. For simple TCP port monitoring you can use tcp://[HOST]:[PORT:80]. It works for add-ons on the host or internal network.
realtimeboolfalseGive add-on access to host schedule including SYS_NICE for change execution time/priority.
journaldboolfalseIf set to true, the host's system journal will be mapped read-only into the add-on. Most of the time the journal will be in /var/log/journal however on some hosts you will find it in /run/log/journal. Add-ons relying on this capability should check if the directory /var/log/journal is populated and fallback on /run/log/journal if not.
breaking_versionslistList of breaking versions of the addon. A manual update will always be required if the update is to a breaking version or would cross a breaking version, even if users have auto-update enabled for the addon.

Options / Schema

The options dictionary contains all available options and their default value. Set the default value to null if the value is required to be given by the user before the add-on can start. Nested arrays and dictionaries are supported with a maximum depth of two. To make an option optional, put ? at the end of the data type, otherwise it will be a required value.

message: "custom things"
- username: beer
password: "123456"
- username: cheep
password: "654321"
- haha
- hihi
link: ""
size: 15
count: 1.2

If you remove a configuration option from an add-on already deployed to users, it is recommended to delete the option to avoid a warning like Option '<options_key>' does not exist in the schema for <Add-on Name> (<add-on slug>).

To remove an option the Supervisor addons API can be used. Using bashio this boils down to bashio::addon.option '<options_key>' (without additional argument to delete this option key). Typically this should be called inside an if block checking if the option is still set using bashio::config.exists '<options_key>'.

The schema looks like options but describes how we should validate the user input. For example:

message: str
- username: str
password: str
- "match(^\\w*$)"
link: url
size: "int(5,20)"
count: float
not_need: "str?"

We support:

  • str / str(min,) / str(,max) / str(min,max)
  • bool
  • int / int(min,) / int(,max) / int(min,max)
  • float / float(min,) / float(,max) / float(min,max)
  • email
  • url
  • password
  • port
  • match(REGEX)
  • list(val1|val2|...)
  • device / device(filter): Device filter can be in the following format: subsystem=TYPE i.e. subsystem=tty for serial devices.

Add-on extended build

Additional build options for an add-on are stored in build.yaml. This file will be read from our build systems. This is only needed if you are not using the default images or need additional things.

armhf: mycustom/base-image:latest
squash: false
my_build_arg: xy
build_fromnoA dictionary with the hardware architecture as the key and the base Docker image as the value.
squashnoDefault False. Be careful with this option, as you can not use the image for caching stuff after that!
argsnoAllow additional Docker build arguments as a dictionary.
labelsnoAllow additional Docker labels as a dictionary.
codenotarynoEnable container signature with codenotary CAS.
codenotary.signernoOwner signer E-Mail address for this image.
codenotary.base_imagenoVerify the base container image. If you use our official images, use [email protected]

We provide a set of base images which should cover a lot of needs. If you don't want to use the Alpine based version or need a specific image tag, feel free to pin this requirement for your build with the build_from option.

Add-on translations

Add-ons can provide translation files for configuration options that are used in the UI.

Example path to translation file: addon/translations/{language_code}.yaml

For {language_code} use a valid language code, like en, for a full list have a look here, en.yaml would be a valid filename.

This file supports 2 main keys configuration and network.

Configuration translations

name: Enable SSL
description: Enable usage of SSL on the webserver inside the add-on

The key under configuration (ssl) in this case, needs to match a key in your schema configuration (in config.yaml).

Port description translations

80/TCP: The webserver port (Not used for Ingress)

The key under network (80/TCP) in this case, needs to match a key in your ports configuration (in config.yaml).

Add-on advanced options

Sometimes add-on developers may want to allow users to configure to provide their own files which are then provided directly to an internal service as part of its configuration. Some examples include:

  1. Internal service wants a list of configured items and the schema of each item is complex but the service provides no UI for doing so, easier to point users to their documentation and ask for a file in that schema.
  2. Internal service requires a binary file or some file configured externally as part of its config.
  3. Internal service supports live reloading on config change and you want to support that for some or all of its configuration by asking users for a file in its schema to live reload from.

In cases like these you should add addon_config to map in your addon's configuration file. And then you should direct your users to put this file in the folder /addon_configs/<your addon's slug>. This folder will be mounted at /config inside your addon's docker container at runtime. You should either provide an option in your addon's schema that collects a relative path to the file(s) starting from this folder or rely on a fixed filename and include that in your documentation.

Another use case of addon_config could be if your addon wants to provide file-based output or give users access to internal files for debugging. Some examples include:

  1. Internal service logs to a file and you wish to allow users access to that log file
  2. Internal service uses a database and you wish to allow users access to that database for debugging
  3. Internal service generates files which are intended to be used in its own config and you wish to allow users to access them as well

In cases like these you should add addon_config:rw to map so your addon can write to this folder as well as read from it. And then you should write these files out to /config during your addon's runtime so users can see and access them.