用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/shreed27/DAIN --skill tts命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | tts |
| description | Text-to-speech synthesis with ElevenLabs and system voices |
| emoji | 🔊 |
| gates | {"envs":{"anyOf":["ELEVENLABS_API_KEY"]}} |
Convert text to natural-sounding speech using ElevenLabs, macOS say, or espeak.
/speak "Your order has been filled" Speak text aloud
/speak "Market alert" --voice rachel Use specific voice
/speak "Portfolio up 5%" --speed 1.2 Adjust speed
/voices List available voices
/voices preview rachel Preview a voice
/voice set rachel Set default voice
/tts status Check TTS status
/tts provider elevenlabs Set provider
/tts speed 1.0 Set default speed
/tts volume 0.8 Set volume (0-1)
import { createTTSService } from 'clodds/tts';
const tts = createTTSService({
provider: ,
: process..,
: ,
: ,
: ,
});
// Basic synthesis
const audio = await tts.synthesize('Hello, your trade was executed.');
// Play immediately
await tts.speak('Portfolio value is $10,000');
// With options
await tts.speak('Market alert: BTC crossed $100k', {
voice: 'josh',
speed: 1.2,
pitch: 1.0,
volume: 0.8,
});
// Stream for long text (lower latency)
const stream = await tts.streamSynthesize(longText, {
voice: 'rachel',
});
stream.on('data', (chunk) => {
// Play audio chunks as they arrive
audioPlayer.write(chunk);
});
stream.on('end', () => {
console.log('Synthesis complete');
});
// Get available voices
const voices = await tts.listVoices();
for (const voice of voices) {
console.log(`${voice.id}: ${voice.name}`);
console.log(` Gender: ${voice.gender}`);
console.log(` Accent: ${voice.accent}`);
console.log(` Use case: ${voice.useCase}`);
}
// Preview a voice
await tts.preview('rachel', 'This is a preview of the Rachel voice.');
// Queue multiple messages
tts.queue('First message');
tts.queue('Second message');
tts.queue('Third message');
// Messages play in order
// Clear queue
tts.clearQueue();
// Skip current
tts.skip();
| Voice ID | Name | Gender | Accent | Best For |
|---|---|---|---|---|
rachel | Rachel | F | American | Narration |
domi | Domi | F | American | Conversational |
bella | Bella | F | American | Soft, gentle |
antoni | Antoni | M | American | Narration |
josh | Josh | M | American | Deep, authoritative |
arnold | Arnold | M | American | Gruff, character |
adam | Adam | M | American | Deep, narration |
sam | Sam | M | American | Raspy, character |
| Provider | Quality | Latency | Cost | Setup |
|---|---|---|---|---|
| ElevenLabs | Premium | ~500ms | $5/100k chars | API key |
| say (macOS) | Good | ~100ms | Free | Built-in |
| espeak | Basic | ~50ms | Free | Install |
// ElevenLabs (best quality)
const tts = createTTSService({
provider: 'elevenlabs',
apiKey: process.env.ELEVENLABS_API_KEY,
});
// macOS say (free, local)
const tts = createTTSService({
provider: 'say',
defaultVoice: 'Samantha', // macOS voice
});
// espeak (cross-platform, free)
const tts = createTTSService({
provider: 'espeak',
defaultVoice: 'en-us',
});
// Set output device
tts.setOutputDevice('Built-in Speakers');
// Get available devices
const devices = await tts.listOutputDevices();
// Use SSML for advanced control
await tts.speak(`
<speak>
<prosody rate="slow">Important alert:</prosody>
<break time="500ms"/>
Your stop loss was triggered.
</speak>
`, { ssml: true });
基于 SOC 职业分类