mirror of
https://github.com/agent0ai/agent-zero.git
synced 2026-05-22 19:47:15 +00:00
675afa8dee
Split the legacy core speech stack into two built-in, independently toggleable plugins: `_kokoro_tts` for TTS and `_whisper_stt` for STT. This refactor keeps dependency installation and bootstrap concerns in Docker/bootstrap/preload, while moving speech-specific tooling, APIs, prompts, UI, and runtime behavior into the plugins. Core now exposes engine-agnostic `tts-service` and `stt-service` brokers, with browser-native TTS preserved as the fallback when Kokoro is disabled. Included in this change: - add built-in `_kokoro_tts` plugin with plugin-owned synth API, config, status UI, and provider registration - add built-in `_whisper_stt` plugin with plugin-owned transcribe API, mic runtime, device UI, prompt injection, and provider registration - remove legacy core speech APIs/helpers/settings/UI and delete unused `webui/js/speech_browser.js` - replace the old hardcoded speech settings section with a generic voice surface backed by plugin extensions - update preload/docs/tests to match the new plugin-owned speech architecture Behavioral intent: - both plugins are built-in but not `always_enabled` - users can now hot-switch TTS and STT independently - browser TTS remains available when `_kokoro_tts` is off - Whisper mic UI only appears when `_whisper_stt` is enabled
223 lines
8.6 KiB
Python
223 lines
8.6 KiB
Python
from __future__ import annotations
|
|
|
|
import importlib
|
|
import sys
|
|
import types
|
|
from pathlib import Path
|
|
|
|
|
|
PROJECT_ROOT = Path(__file__).resolve().parents[1]
|
|
if str(PROJECT_ROOT) not in sys.path:
|
|
sys.path.insert(0, str(PROJECT_ROOT))
|
|
|
|
|
|
from helpers import plugins, settings
|
|
|
|
|
|
def test_builtin_speech_plugins_are_discoverable_and_toggleable() -> None:
|
|
discovered = {
|
|
item.name: item
|
|
for item in plugins.get_enhanced_plugins_list(
|
|
custom=True,
|
|
builtin=True,
|
|
plugin_names=["_kokoro_tts", "_whisper_stt"],
|
|
)
|
|
}
|
|
|
|
assert "_kokoro_tts" in discovered
|
|
assert "_whisper_stt" in discovered
|
|
|
|
assert discovered["_kokoro_tts"].always_enabled is False
|
|
assert discovered["_whisper_stt"].always_enabled is False
|
|
assert "agent" in discovered["_kokoro_tts"].settings_sections
|
|
assert "agent" in discovered["_whisper_stt"].settings_sections
|
|
|
|
|
|
def test_legacy_core_speech_artifacts_are_removed() -> None:
|
|
removed_paths = [
|
|
"api/synthesize.py",
|
|
"api/transcribe.py",
|
|
"helpers/kokoro_tts.py",
|
|
"helpers/whisper.py",
|
|
"webui/components/chat/speech/speech-store.js",
|
|
"webui/components/settings/agent/speech.html",
|
|
"webui/components/settings/speech/microphone-setting-store.js",
|
|
"webui/components/settings/speech/microphone.html",
|
|
"webui/css/speech.css",
|
|
"webui/js/speech_browser.js",
|
|
]
|
|
|
|
for relative_path in removed_paths:
|
|
assert not (PROJECT_ROOT / relative_path).exists(), relative_path
|
|
|
|
|
|
def test_plugin_owned_voice_files_exist() -> None:
|
|
expected_paths = [
|
|
"plugins/_kokoro_tts/plugin.yaml",
|
|
"plugins/_kokoro_tts/api/synthesize.py",
|
|
"plugins/_kokoro_tts/extensions/webui/page-head/runtime.html",
|
|
"plugins/_kokoro_tts/extensions/webui/voice-settings-main/kokoro-card.html",
|
|
"plugins/_whisper_stt/plugin.yaml",
|
|
"plugins/_whisper_stt/api/transcribe.py",
|
|
"plugins/_whisper_stt/extensions/webui/page-head/runtime.html",
|
|
"plugins/_whisper_stt/extensions/webui/chat-input-box-end/microphone-button.html",
|
|
"plugins/_whisper_stt/extensions/webui/voice-settings-main/whisper-card.html",
|
|
"plugins/_whisper_stt/webui/whisper-stt-store.js",
|
|
]
|
|
|
|
for relative_path in expected_paths:
|
|
assert (PROJECT_ROOT / relative_path).exists(), relative_path
|
|
|
|
|
|
def test_core_settings_no_longer_expose_legacy_speech_keys() -> None:
|
|
defaults = settings.get_default_settings()
|
|
output = settings.convert_out(defaults)
|
|
|
|
legacy_keys = {
|
|
"tts_kokoro",
|
|
"stt_model_size",
|
|
"stt_language",
|
|
"stt_silence_threshold",
|
|
"stt_silence_duration",
|
|
"stt_waiting_timeout",
|
|
}
|
|
|
|
assert legacy_keys.isdisjoint(defaults.keys())
|
|
assert legacy_keys.isdisjoint(output["settings"].keys())
|
|
assert "stt_models" not in output["additional"]
|
|
|
|
|
|
def test_voice_prefix_prompt_rule_is_removed() -> None:
|
|
core_prompt = (PROJECT_ROOT / "prompts/agent.system.main.communication_additions.md").read_text(
|
|
encoding="utf-8"
|
|
)
|
|
whisper_store = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/webui/whisper-stt-store.js"
|
|
).read_text(encoding="utf-8")
|
|
voice_surface = (PROJECT_ROOT / "webui/components/settings/agent/voice.html").read_text(
|
|
encoding="utf-8"
|
|
)
|
|
|
|
assert "if starts (voice) then transcribed can contain errors consider compensation" not in core_prompt
|
|
assert "(voice)" not in whisper_store
|
|
assert not (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/prompts/agent.system.voice_transcription.md"
|
|
).exists()
|
|
assert not (
|
|
PROJECT_ROOT
|
|
/ "plugins/_whisper_stt/extensions/python/system_prompt/_20_voice_transcription.py"
|
|
).exists()
|
|
assert '<x-extension id="voice-settings-start"></x-extension>' in voice_surface
|
|
assert '<x-extension id="voice-settings-main"></x-extension>' in voice_surface
|
|
assert '<x-extension id="voice-settings-end"></x-extension>' in voice_surface
|
|
|
|
|
|
def test_whisper_message_mode_defaults_to_send_and_supports_draft() -> None:
|
|
sys.modules.setdefault(
|
|
"whisper",
|
|
types.SimpleNamespace(load_model=lambda *args, **kwargs: None),
|
|
)
|
|
runtime = importlib.import_module("plugins._whisper_stt.helpers.runtime")
|
|
|
|
assert runtime.normalize_config({})["message_mode"] == "send"
|
|
assert runtime.normalize_config({"message_mode": "draft"})["message_mode"] == "draft"
|
|
assert runtime.normalize_config({"message_mode": "DRAFT"})["message_mode"] == "draft"
|
|
assert runtime.normalize_config({"message_mode": "invalid"})["message_mode"] == "send"
|
|
|
|
default_config = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/default_config.yaml"
|
|
).read_text(encoding="utf-8")
|
|
migration = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/helpers/migration.py"
|
|
).read_text(encoding="utf-8")
|
|
config_ui = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/webui/config.html"
|
|
).read_text(encoding="utf-8")
|
|
status_ui = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/webui/main.html"
|
|
).read_text(encoding="utf-8")
|
|
voice_card = (
|
|
PROJECT_ROOT
|
|
/ "plugins/_whisper_stt/extensions/webui/voice-settings-main/whisper-card.html"
|
|
).read_text(encoding="utf-8")
|
|
whisper_store = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/webui/whisper-stt-store.js"
|
|
).read_text(encoding="utf-8")
|
|
|
|
assert "message_mode: send" in default_config
|
|
assert '"message_mode": "send"' in migration
|
|
assert '<option value="send">Send immediately</option>' in config_ui
|
|
assert '<option value="draft">Draft in composer</option>' in config_ui
|
|
assert "messageModeLabel" in status_ui
|
|
assert "messageModeLabel" in voice_card
|
|
assert 'message_mode: "send"' in whisper_store
|
|
assert 'status?.config?.message_mode === "draft" ? "draft" : "send"' in whisper_store
|
|
assert "updateChatInput(message)" in whisper_store
|
|
assert "sendMessage()" in whisper_store
|
|
|
|
|
|
def test_browser_tool_speech_action_uses_shared_tts_service() -> None:
|
|
browser_handler = (
|
|
PROJECT_ROOT
|
|
/ "plugins/_browser/extensions/webui/get_tool_message_handler/browser-tool-handler.js"
|
|
).read_text(encoding="utf-8")
|
|
|
|
assert "/components/chat/speech/speech-store.js" not in browser_handler
|
|
assert "/js/tts-service.js" in browser_handler
|
|
assert "ttsService.speak(contentText)" in browser_handler
|
|
|
|
|
|
def test_chat_bar_keeps_existing_send_and_mic_icon_contract() -> None:
|
|
chat_bar = (
|
|
PROJECT_ROOT / "webui/components/chat/input/chat-bar-input.html"
|
|
).read_text(encoding="utf-8")
|
|
mic_extension = (
|
|
PROJECT_ROOT
|
|
/ "plugins/_whisper_stt/extensions/webui/chat-input-box-end/microphone-button.html"
|
|
).read_text(encoding="utf-8")
|
|
whisper_store = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/webui/whisper-stt-store.js"
|
|
).read_text(encoding="utf-8")
|
|
whisper_css = (
|
|
PROJECT_ROOT / "plugins/_whisper_stt/webui/whisper-stt.css"
|
|
).read_text(encoding="utf-8")
|
|
|
|
assert 'id="send-button"' in chat_bar
|
|
assert 'x-text="$store.chatInput.sendButtonIcon"' in chat_bar
|
|
assert ':class="$store.chatInput.sendButtonClass"' in chat_bar
|
|
assert ':title="$store.chatInput.sendButtonTitle"' in chat_bar
|
|
|
|
assert 'id="microphone-button"' in mic_extension
|
|
assert "<svg" in mic_extension
|
|
assert "material-symbols-outlined" not in mic_extension
|
|
assert "buttonIcon" not in mic_extension
|
|
assert 'title=' not in mic_extension
|
|
assert 'x-effect="$store.whisperStt.updateMicrophoneButtonUI()"' in mic_extension
|
|
assert 'x-init="$store.whisperStt.updateMicrophoneButtonUI()"' in mic_extension
|
|
assert "updateMicrophoneButtonUI()" in whisper_store
|
|
assert "data-status" in whisper_store
|
|
assert 'setAttribute("title"' not in whisper_store
|
|
assert 'removeAttribute("title")' in whisper_store
|
|
assert 'removeAttribute("data-bs-original-title")' in whisper_store
|
|
assert "this.updateMicrophoneButtonUI();" in whisper_store
|
|
assert "sttService.emitStatusChange(this.micStatus)" in whisper_store
|
|
|
|
for state in [
|
|
"disabled",
|
|
"inactive",
|
|
"activating",
|
|
"listening",
|
|
"recording",
|
|
"waiting",
|
|
"processing",
|
|
]:
|
|
assert f'"mic-{state}"' in whisper_store
|
|
assert f"#microphone-button.mic-{state}" in whisper_css
|
|
assert f"#microphone-button.mic-{state}" in mic_extension
|
|
|
|
assert "background-color: red;" in whisper_css
|
|
assert "background-color: green;" in whisper_css
|
|
assert "background-color: teal;" in whisper_css
|
|
assert "background-color: grey;" in mic_extension
|
|
assert "whisper-stt-mic-pulse 0.8s infinite" in mic_extension
|
|
assert "whisper-stt-mic-pulse 0.8s infinite" in whisper_css
|