Florian Asche
3b33730105
* feat(ai): integrate Kilo agent framework and coding instructions Add configuration files, agent definitions, and skill sets to enable AI-driven development workflows and standardized agent behavior.
129 lines
5.0 KiB
Markdown
129 lines
5.0 KiB
Markdown
# Project Context for AI Agents
|
|
|
|
## Behaviour
|
|
|
|
- NEVER automatically reply on Github (PRs or Discussions) without explicit consent from the developer.
|
|
- Always run `./script/lint` and `./script/tests` after code changes to ensure standards are met.
|
|
|
|
## Project Overview
|
|
|
|
**Linux Voice Assistant (LVA)** - A Linux-based voice satellite software for Home Assistant using the ESPHome protocol.
|
|
|
|
### Purpose
|
|
- Provides voice control and interaction for Home Assistant via the Assist integration
|
|
- Runs on x64/ARM64 Linux hardware with PulseAudio/PipeWire audio support
|
|
- Performs local wake word detection (OpenWakeWord/MicroWakeWord) and on-device processing
|
|
|
|
### Architecture
|
|
- **Protocol**: ESPHome API via `aioesphomeapi` for Home Assistant integration
|
|
- **Wake Words**: Supports both OpenWakeWord and MicroWakeWord engines
|
|
- **Audio**: Uses `soundcard` library for input, `python-mpv` for output
|
|
- **Peripherals**: WebSocket API for LEDs, buttons, HAT boards
|
|
- **Discovery**: mDNS/ZeroConf for auto-discovery by Home Assistant
|
|
|
|
### Key Components
|
|
- `linux_voice_assistant/__main__.py` - Main entry point with CLI argument parsing
|
|
- `linux_voice_assistant/satellite.py` - VoiceSatelliteProtocol implementation (ESPHome API)
|
|
- `linux_voice_assistant/wake_word.py` - Wake word model loading and detection
|
|
- `linux_voice_assistant/webrtc.py` - WebRTC noise gain/NS processing
|
|
- `linux_voice_assistant/mpv_player.py` - MPV-based media player
|
|
- `linux_voice_assistant/entity.py` - Entity definitions (MediaPlayer, MuteSwitch, etc.)
|
|
- `linux_voice_assistant/api_server.py` - Peripheral WebSocket API server
|
|
- `linux_voice_assistant/models.py` - Preferences and ServerState dataclasses
|
|
|
|
## Development Commands
|
|
|
|
### Setup
|
|
```bash
|
|
./script/setup --dev # Install dev dependencies in virtual environment
|
|
```
|
|
|
|
### Linting
|
|
```bash
|
|
./script/lint # Run all linting checks
|
|
./script/lint_black # Black formatting check (add --auto to fix)
|
|
./script/lint_isort # isort import sorting check (add --auto to fix)
|
|
./script/lint_flake8 # flake8 style checks
|
|
./script/lint_mypy # mypy type checking
|
|
./script/lint_pylint # pylint code quality checks
|
|
```
|
|
|
|
### Testing
|
|
```bash
|
|
./script/tests # Run pytest unit tests
|
|
```
|
|
|
|
## Code Quality Standards
|
|
|
|
- **Python**: 3.11, 3.12, 3.13 supported
|
|
- **Formatting**: Black (200 char line length, py312/py312 target)
|
|
- **Import Sorting**: isort with black profile
|
|
- **Type Checking**: mypy with strict settings
|
|
- **Linting**: pylint (many checks disabled in pyproject.toml for practical reasons)
|
|
|
|
## Testing Strategy
|
|
|
|
- Unit tests in `tests/unit/` using pytest with asyncio mode
|
|
- Test helper `make_state()` in `tests/unit/conftest.py` for creating state fixtures
|
|
- Mocks used heavily for hardware dependencies (audio devices, MPV, wake words)
|
|
|
|
## CI/CD
|
|
|
|
- GitHub Actions for lint checks and unit tests
|
|
- Tests run on Python 3.12 with libmpv-dev system dependency
|
|
- Lint runs on Python 3.13
|
|
|
|
## Branching and PRs
|
|
|
|
- All PRs target `main` (primary development branch). `stable` is for production releases.
|
|
- PRs labeled `bugfix` + `backport-to-stable` are automatically backported to `stable` — use only for bugs also present in `stable`.
|
|
|
|
## Agent Commands
|
|
|
|
When making code changes, run these commands in order:
|
|
|
|
1. **Format code**: `./script/lint_black --auto` and `./script/lint_isort --auto`
|
|
2. **Run linting**: `./script/lint`
|
|
3. **Run tests**: `./script/tests`
|
|
|
|
## Code Style
|
|
|
|
### Comments
|
|
|
|
Only use comments to explain complex, multi-line blocks of code. Do not comment obvious operations. Inline comments are there to explain code parts that need explaining - keep them when editing code.
|
|
|
|
### Docstring Format
|
|
|
|
Use Sphinx-style docstrings with `:param:` syntax. For simple functions, a single-line docstring is fine. The docstring should provide clarity to the caller, not explain how it works technically.
|
|
|
|
```python
|
|
def my_function(param1: str, param2: int, param3: bool = False) -> str:
|
|
"""
|
|
Brief one-line description of the function.
|
|
:param param1: Description of what param1 is used for.
|
|
:param param2: Description of what param2 is used for.
|
|
:param param3: Description of what param3 is used for.
|
|
"""
|
|
```
|
|
|
|
Do **not** use Google-style (`Args:`) or bullet-style (`- param:`) docstrings.
|
|
|
|
### File structure
|
|
|
|
Private methods should be at the bottom of the file, public at the top.
|
|
|
|
## Verification Checklist
|
|
|
|
Before claiming completion:
|
|
- [ ] Ran `./script/lint` - all checks passed
|
|
- [ ] Ran `./script/tests` - all tests passed
|
|
- [ ] For audio-related changes: Note that hardware testing (microphone/speaker) was NOT performed
|
|
- [ ] Did NOT claim hardware behavior is verified unless actually exercised
|
|
- [ ] Changes are focused on one feature or fix per change
|
|
|
|
## Areas Requiring Extra Care
|
|
|
|
- `linux_voice_assistant/wake_word.py` - Wake word model loading is hardware-dependent
|
|
- `linux_voice_assistant/mpv_player.py` - Audio output requires `libmpv-dev`
|
|
- `linux_voice_assistant/webrtc.py` - Noise suppression/gain algorithms
|
|
- `linux_voice_assistant/satellite.py` - ESPHome API protocol handling |