HassIL
The Home Assistant Intent Language (HassIL) parser for intents.
Dependencies
- PyYAML
Installation
Run the script/setup script to automatically create a virtual environment and install the requirements.
Running
python3 -m hassil <yaml_file_or_directory> [<yaml_file_or_directory> ...]
Once loaded, you may type in a sentence and see what intent it matches. For example:
python3 -m hassil examples/en.yaml --areas 'living room'
what is the temperature in the living room
{'intent': 'HassClimateGetTemperature', 'area': 'living room', 'domain': 'climate'}
Make sure to provide area names with --areas. Device or entity names can be provided with --names.
python3 -m hassil examples/en.yaml --areas office --names trapdoor
open the trapdoor in the office
{'intent': 'HassOpenCover', 'name': 'trapdoor', 'area': 'office'}
Sampling Sentences
Sentences for each intent can be sampled from the intent YAML files:
python3 -m hassil.sample examples/en.yaml -n 1
{"intent": "HassTurnOn", "text": "turn on the entity"}
{"intent": "HassTurnOff", "text": "turn off the entity"}
{"intent": "HassOpenCover", "text": "open the entity in the area"}
{"intent": "HassCloseCover", "text": "close the entity in the area"}
{"intent": "HassLightsSet", "text": "set the entity color to red"}
{"intent": "HassClimateSetTemperature", "text": "set temperature to 0 degrees in the area"}
{"intent": "HassClimateGetTemperature", "text": "what is the temperature in the area"}
The --areas and --names arguments are the same from python3 -m hassil, but default to generic "area" and "entity" terms.
Exclude the -n argument to sample all possible sentences.
Sentence Templates
Uses a custom parser written in Python.
- Alternative words or phrases
(red|green|blue)turn(s|ed|ing)
- Optional words or phrases
[the][this|that]light[s]
- Permutations of words or phrases
(patience; you must have) my young Padawanis [the] light (on; in <area>)
- Slot Lists
{list_name}{list_name:slot_name}- Refers to a pre-defined list of values in YAML (
lists), either global or local (particular to the intent to which the sentence refers)
- Expansion Rules
<rule_name>- Refers to a pre-defined expansion rule in YAML (
expansion_rules), either global or local (particular to the intent to which the sentence refers)
YAML Format
language: "<language code>"
intents:
<intent name>:
data:
# List of sentences/slots/etc.
- sentences:
- "<sentence template>"
- "<sentence template>"
# Optional
slots:
# Fixed slots for the recognized intent
<name>: <value>
requires_context:
# Must be present in match context
<name>: # Any provided value is good
excludes_context:
# Must NOT be present in match context
<name>: <value or list>
expansion_rules:
# Expansion rules which only apply to the intent, referenced as <rule_name>
<rule_name>: <sentence template>
lists:
# Lists which apply only to the current set of sentences, referenced as {list_name} or {list_name:slot_name}
<list name>:
values:
# See below for other possible types
- "items"
- "in list"
metadata:
# Arbitrary key/value pairs that will be available in the result
<key>: <value>
# Optional lists of items that become alternatives in sentence templates
lists:
# Referenced as {list_name} or {list_name:slot_name}
<list name>:
values:
- "items"
- "in list"
- in: "text in"
out: <value for slot>
# Optional
context:
<name>: <value>
metadata:
# Arbitrary key/value pairs that will be available in the result
<key>: <value>
<range_name>
range:
type: "number"
from: 0
to: 100 # inclusive
multiplier: 1.0 # multiply to get final value
<wildcard_name>
wildcard: true
# Optional rules that are expanded in sentence templates
expansion_rules:
# Referenced as <rule_name>
<rule_name>: "<sentence template>"
# Optional words that the intent recognizer can skip during recognition
skip_words:
- "<word>"
Inline Range Lists
Within a sentence template, you can create a range list inline with the format {start..end,[step]:slot_name}. For example:
language: "en"
intents:
SetBrightness:
data:
- sentences:
- "set brightness to {0..100:brightness}"
During matching, a range list from 0 to 100 will be available and its value will be stored in the brightness slot.
An optional step can be used as well:
language: "en"
intents:
StartTimer:
data:
- sentences:
- "start timer for {10..100,10:seconds} seconds"
This range list goes from 10 to 100 in steps of 10, so the sentence "start timer for 11 seconds" will not be recognized.
2 maintainers
Release files for hassil 3.12.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| hassil-3.12.0.tar.gz | 69.7 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| hassil-3.12.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 127.0 kB
Release files / hassil-3.12.0.tar.gz
| Download URL | hassil-3.12.0.tar.gz |
|---|---|
| Size | 69.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
848397674400036d64bf527bf2f142944c07e03330bdcae1ee015655c03cfcc5
|
|
BLAKE2b-256 checksum How to use checksums |
d91fc9494eb84fa7a901bcb1316e8196789400d189841decf0b5854f87f0aec3
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Aug 25, 2026.
Transparency logRelease files / hassil-3.12.0-py3-none-any.whl
| Download URL | hassil-3.12.0-py3-none-any.whl |
|---|---|
| Size | 57.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
4f5fed98753c9f730c029a1142b79d89d17305c752ba27238e8cdcb7ac1f6e17
|
|
BLAKE2b-256 checksum How to use checksums |
d5ed90a008f651e16fee57f8167a1b897b024ead11156d318af217a01d1c1ea1
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Aug 25, 2026.
Transparency log
