axterminator 0.10.1

macOS GUI testing framework with background testing, sub-millisecond element access, and self-healing locators
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75

# Audio & Speech

AXTerminator includes on-device audio capture, speech recognition, and text-to-speech via the `audio` feature flag.

## Enable Audio

Build with the `audio` feature:

```bash
cargo build --release --features "cli,audio"
```

## MCP Tools

| Tool | Description |
|------|-------------|
| `ax_listen` | Capture microphone or system audio, transcribe via SFSpeechRecognizer |
| `ax_speak` | Text-to-speech via system, Kokoro, or Piper engines |
| `ax_audio_voices` | List installed macOS speech voices |
| `ax_audio_devices` | List available audio input/output devices |

## Speech Recognition

AXTerminator captures audio at **native 48kHz sample rate** and transcribes on-device using Apple's SFSpeechRecognizer. No cloud API required.

### Requirements

- **Dictation** must be enabled: System Settings > Keyboard > Dictation
- **Microphone permission** granted to your terminal app

### How It Works

1. CoreAudio captures from the default input device at 48kHz
2. Audio is buffered and written as a WAV file
3. SFSpeechRecognizer transcribes on-device
4. CFRunLoop is pumped for callback delivery

!!! note "48kHz Native Capture"
    Earlier versions captured at 48kHz but wrote 16kHz WAV headers, causing garbled audio. v0.6.0 fixed this to write correct 48kHz headers, resolving speech recognition accuracy.

## Text-to-Speech

`ax_speak` defaults to the existing macOS `NSSpeechSynthesizer` path:

```json
{"text": "Hello from AXTerminator"}
```

To opt into local neural engines, build with `enhanced-tts`, install the
external `sherpa-onnx-offline-tts` runtime, and download model files on demand:

```bash
cargo build --release --features "cli,enhanced-tts"
axterminator models tts list
axterminator models tts download kokoro
axterminator models tts download piper
```

Then call:

```json
{"text": "Hello from AXTerminator", "engine": "kokoro", "voice": "af_heart"}
```

If an enhanced engine is requested but its model files are missing, `ax_speak`
falls back to `system` and returns `fallback_reason`.

## Audio Devices

```bash
# List input/output devices
axterminator audio-devices
```

Lists all CoreAudio devices with their sample rates and channel counts.