@hasna/skills 0.0.2

package/skills/skill-generate-audio/src/index.ts

@@ -0,0 +1,656 @@
+#!/usr/bin/env bun
+/**
+ * Text to Speech Skill
+ * Converts text to natural speech using multiple AI providers:
+ * - ElevenLabs (v3, v2, turbo, flash)
+ * - OpenAI TTS (tts-1-hd, tts-1)
+ * - Google Cloud TTS (optional)
+ *
+ * API Documentation:
+ * - ElevenLabs: https://elevenlabs.io/docs/api-reference/text-to-speech
+ * - OpenAI: https://platform.openai.com/docs/guides/text-to-speech
+ */
+
+import { parseArgs } from "util";
+import { mkdirSync, writeFileSync, appendFileSync, existsSync, readFileSync } from "fs";
+import { join, dirname } from "path";
+import { randomUUID } from "crypto";
+
+// ============================================================================ 
+// Types
+// ============================================================================ 
+type Provider = "elevenlabs" | "openai";
+
+interface TTSOptions {
+  text: string;
+  provider: Provider;
+  voice: string;
+  model: string;
+  format: string;
+  quality: string;
+  stability: number;
+  clarity: number;
+  style: number;
+  speed: number;
+  language?: string;
+}
+
+interface VoiceSettings {
+  stability: number;
+  similarity_boost: number;
+  style: number;
+  use_speaker_boost: boolean;
+  speed: number;
+}
+
+// ============================================================================ 
+// Constants
+// ============================================================================ 
+const SKILL_NAME = "generate-audio";
+const SESSION_ID = randomUUID().slice(0, 8);
+
+// Use SKILLS_OUTPUT_DIR from CLI if available
+const SKILLS_OUTPUT_DIR = process.env.SKILLS_OUTPUT_DIR || join(process.cwd(), ".skills");
+const EXPORTS_DIR = join(SKILLS_OUTPUT_DIR, "exports", SKILL_NAME);
+const LOGS_DIR = join(SKILLS_OUTPUT_DIR, "logs", SKILL_NAME);
+
+// Session timestamp for log filename
+const SESSION_TIMESTAMP = new Date()
+  .toISOString()
+  .replace(/[:.]/g, "_")
+  .replace(/-/g, "_")
+  .slice(0, 19)
+  .toLowerCase();
+
+// ============================================================================ 
+// Provider Configurations
+// ============================================================================ 
+const PROVIDERS: Record<Provider, { name: string; envKey: string }> = {
+  elevenlabs: { name: "ElevenLabs", envKey: "ELEVENLABS_API_KEY" },
+  openai: { name: "OpenAI TTS", envKey: "OPENAI_API_KEY" },
+};
+
+// ============================================================================ 
+// ElevenLabs Voice IDs (Default Premium Voices)
+// ============================================================================ 
+const ELEVENLABS_VOICE_IDS: Record<string, string> = {
+  // Female voices - Narrative & Professional
+  rachel: "21m00Tcm4TlvDq8ikWAM", // Warm, professional - narration, podcasts
+  domi: "AZnzlk1XvdvUeBnXmlld", // Energetic - marketing, ads
+  bella: "EXAVITQu4vr4xnSDxMaL", // Soft, gentle
+  elli: "MF3mGyEYCl7XYWbV9V6O", // Young female
+  charlotte: "XB0fDUnXU5powFXDhCwa", // Swedish female
+  matilda: "XrExE9yKIg1WjnnlVkGX", // Warm - children's content
+  dorothy: "ThT5KcBeYPX3keUQqHPh", // Pleasant British
+  sarah: "EXAVITQu4vr4xnSDxMaL", // Soft, soothing
+  emily: "LcfcDJNUP1GQjkzn1xUU", // American female
+  gigi: "jBpfuIE2acCO8z3wKNLl", // Expressive
+  freya: "jsCqWAovK2LkecY7zXl4", // Warm
+  grace: "oWAxZDx7w5VEj9dCyTzz", // Professional
+  lily: "pFZP5JQG7iQjIQuC4Bku", // British, elegant
+  serena: "pMsXgVXv3BLzUgSXRplE", // Soothing
+  nicole: "piTKgcLEGmPE4e6mEKli", // Clear
+  jessie: "t0jbNlBVZ17f02VDIeMI", // Friendly
+  glinda: "z9fAnlkpzviPz146aGWa", // Magical
+  mimi: "zrHiDhphv9ZnVXBqCLjz", // Swedish, gentle
+
+  // Male voices - Narrative & Professional
+  drew: "29vD33N1CtxCmqQRPOHJ", // Authoritative - documentaries
+  clyde: "2EiwWnXFnvU5JabPnv8n", // Deep, resonant - audiobooks
+  paul: "5Q0t7uMcjvnagumLfvZi", // Friendly conversational
+  dave: "CYw3kZ02Hs0563khs1Fj", // Casual British
+  fin: "D38z5RcWu1voky8WS1ja", // Irish
+  antoni: "ErXwobaYiN019PkySvjV", // Young male
+  thomas: "GBv7mTt0atIp3Br8iCZE", // Calm - educational
+  charlie: "IKne3meq5aSn9XLyUdCD", // Australian
+  george: "JBFqnCBsd6RMkjVDRZzb", // British narrator - audiobooks
+  callum: "N2lVS1w4EtoT3dr4eOWO", // Scottish
+  patrick: "ODq5zmih8GrVes37Dizd", // Deep - trailers
+  harry: "SOYHLrjzK2X1ezoPC6cr", // Young British
+  liam: "TX3LPaxmHKxFdv7VOQHJ", // Irish - storytelling
+  josh: "TxGEqnHWrfWFTfGW9XjX", // Young American
+  arnold: "VR6AewLTigWG4xSOukaG", // Robust - action
+  matthew: "Yko7PKHZNXotIFUBG7I9", // British - formal
+  james: "ZQe5CZNOzWyzPSCn5a3c", // Australian - news
+  joseph: "Zlb1dXrM653N07WRdFW3", // British - classic
+  jeremy: "bVMeCyTHy58xNoL34h3p", // American conversational
+  michael: "flq6f7yk4E4fJM5XTYuZ", // American business
+  ethan: "g5CIjZEefAph4nQFvHAz", // American young adult
+  chris: "iP95p4xoKVk53GoZ742B", // Energetic
+  daniel: "onwK4e9ZLuTAKqWW03F9", // Deep British - authority
+  adam: "pNInz6obpgDQGcFmaJgB", // Deep American - professional
+  bill: "pqHfZKP75CvOlQylNhV4", // Trustworthy
+  sam: "yoZ06aMxZJJ28mfd3POQ", // Neutral
+  giovanni: "zcAOhNBS3c14rBihAFp1", // Italian - expressive
+};
+
+// ============================================================================ 
+// OpenAI Voice Names
+// ============================================================================ 
+const OPENAI_VOICES = ["alloy", "echo", "fable", "onyx", "nova", "shimmer"];
+
+// ============================================================================ 
+// Model Configurations
+// ============================================================================ 
+const ELEVENLABS_MODELS: Record<string, { id: string; description: string; supportsAudioTags: boolean }> = {
+  v3: {
+    id: "eleven_v3",
+    description: "Eleven v3 (Alpha) - Most expressive, supports audio tags",
+    supportsAudioTags: true,
+  },
+  v2: {
+    id: "eleven_multilingual_v2",
+    description: "Multilingual v2 - Highest quality, 29 languages",
+    supportsAudioTags: false,
+  },
+  multilingual: {
+    id: "eleven_multilingual_v2",
+    description: "Multilingual v2 - Highest quality, 29 languages",
+    supportsAudioTags: false,
+  },
+  turbo: {
+    id: "eleven_turbo_v2_5",
+    description: "Turbo v2.5 - Low latency (~300ms), English optimized",
+    supportsAudioTags: false,
+  },
+  flash: {
+    id: "eleven_flash_v2_5",
+    description: "Flash v2.5 - Ultra-low latency (75ms), real-time applications",
+    supportsAudioTags: false,
+  },
+  v1: {
+    id: "eleven_monolingual_v1",
+    description: "Monolingual v1 - Legacy English model",
+    supportsAudioTags: false,
+  },
+};
+
+const OPENAI_MODELS: Record<string, { id: string; description: string }> = {
+  "tts-1-hd": {
+    id: "tts-1-hd",
+    description: "OpenAI TTS HD - Highest quality",
+  },
+  "tts-1": {
+    id: "tts-1",
+    description: "OpenAI TTS - Fast, good quality",
+  },
+};
+
+// ============================================================================ 
+// Output Format Configurations
+// ============================================================================ 
+const ELEVENLABS_FORMATS: Record<string, Record<string, string>> = {
+  mp3: {
+    low: "mp3_22050_32",
+    medium: "mp3_44100_64",
+    high: "mp3_44100_128",
+    ultra: "mp3_44100_192",
+  },
+  wav: {
+    low: "pcm_16000",
+    medium: "pcm_22050",
+    high: "pcm_44100",
+    ultra: "pcm_48000",
+  },
+  ogg: {
+    low: "opus_48000_32",
+    medium: "opus_48000_64",
+    high: "opus_48000_128",
+    ultra: "opus_48000_192",
+  },
+  flac: {
+    low: "pcm_22050",
+    medium: "pcm_44100",
+    high: "pcm_44100",
+    ultra: "pcm_48000",
+  },
+};
+
+// OpenAI supported formats
+const OPENAI_FORMATS = ["mp3", "opus", "aac", "flac", "wav", "pcm"];
+
+// File extensions
+const FORMAT_EXTENSIONS: Record<string, string> = {
+  mp3: "mp3",
+  wav: "wav",
+  ogg: "ogg",
+  flac: "flac",
+  opus: "opus",
+  aac: "aac",
+  pcm: "pcm",
+};
+
+// ============================================================================ 
+// Utility Functions
+// ============================================================================ 
+function ensureDir(dir: string) {
+  if (!existsSync(dir)) {
+    mkdirSync(dir, { recursive: true });
+  }
+}
+
+function log(message: string, level: "info" | "error" | "success" | "warn" = "info") {
+  const timestamp = new Date().toISOString();
+  const logFile = join(LOGS_DIR, `log_${SESSION_TIMESTAMP}_${SESSION_ID}.log`);
+
+  ensureDir(LOGS_DIR);
+
+  const logEntry = `[${timestamp}] [${level.toUpperCase()}] ${message}\n`;
+  appendFileSync(logFile, logEntry);
+
+  const prefix = level === "error" ? "❌" : level === "success" ? "✅" : level === "warn" ? "⚠️" : "ℹ️";
+  console.log(`${prefix} ${message}`);
+}
+
+function slugify(text: string): string {
+  return text
+    .toLowerCase()
+    .replace(/\\\[.*?\\\\]/g, "") // Remove audio tags for slug
+    .replace(/[^a-z0-9]+/g, "_")
+    .replace(/^_|_$/g, "")
+    .slice(0, 30);
+}
+
+function detectAudioTags(text: string): string[] {
+  const tagRegex = /\\\\[[^\\]+\\\\]/g;
+  const matches = text.match(tagRegex);
+  return matches || [];
+}
+
+function getProvider(model: string): Provider {
+  if (model.startsWith("tts-")) {
+    return "openai";
+  }
+  return "elevenlabs";
+}
+
+// ============================================================================ 
+// ElevenLabs API Integration
+// ============================================================================ 
+async function generateWithElevenLabs(options: TTSOptions): Promise<ArrayBuffer> {
+  const apiKey = process.env.ELEVENLABS_API_KEY || process.env.ELEVEN_API_KEY;
+
+  if (!apiKey) {
+    throw new Error("ELEVENLABS_API_KEY environment variable is not set");
+  }
+
+  // Resolve voice ID
+  const voiceId = ELEVENLABS_VOICE_IDS[options.voice.toLowerCase()] || options.voice;
+
+  // Resolve model
+  const modelKey = options.model.toLowerCase();
+  const modelConfig = ELEVENLABS_MODELS[modelKey] || ELEVENLABS_MODELS.v2;
+  const modelId = modelConfig.id;
+
+  // Check for audio tags in text
+  const audioTags = detectAudioTags(options.text);
+  if (audioTags.length > 0) {
+    log(`Detected audio tags: ${audioTags.join(", ")}`);
+    if (!modelConfig.supportsAudioTags) {
+      log(`Warning: Model '${modelKey}' does not support audio tags. Consider using 'v3' for best results.`, "warn");
+    }
+  }
+
+  // Resolve output format
+  const formatConfig = ELEVENLABS_FORMATS[options.format.toLowerCase()] || ELEVENLABS_FORMATS.mp3;
+  const outputFormat = formatConfig[options.quality.toLowerCase()] || formatConfig.high;
+
+  log(`Provider: ElevenLabs`);
+  log(`Voice: ${options.voice} (${voiceId})`);
+  log(`Model: ${modelId} - ${modelConfig.description}`);
+  log(`Format: ${outputFormat}`);
+  log(`Text length: ${options.text.length} characters`);
+
+  // Build voice settings
+  const voiceSettings: VoiceSettings = {
+    stability: options.stability,
+    similarity_boost: options.clarity,
+    style: options.style,
+    use_speaker_boost: true,
+    speed: options.speed,
+  };
+
+  log(
+    `Voice settings: stability=${options.stability}, clarity=${options.clarity}, style=${options.style}, speed=${options.speed}`
+  );
+
+  // Build request body
+  const body: Record<string, unknown> = {
+    text: options.text,
+    model_id: modelId,
+    voice_settings: voiceSettings,
+  };
+
+  // Add language if specified
+  if (options.language) {
+    body.language_code = options.language;
+    log(`Language: ${options.language}`);
+  }
+
+  // Make API request
+  const url = `https://api.elevenlabs.io/v1/text-to-speech/${voiceId}?output_format=${outputFormat}`;
+
+  const response = await fetch(url, {
+    method: "POST",
+    headers: {
+      "Content-Type": "application/json",
+      "xi-api-key": apiKey,
+    },
+    body: JSON.stringify(body),
+  });
+
+  if (!response.ok) {
+    const errorText = await response.text();
+    let errorMessage = `ElevenLabs API error: ${response.status} ${response.statusText}`;
+
+    try {
+      const errorJson = JSON.parse(errorText);
+      if (errorJson.detail?.message) {
+        errorMessage = `ElevenLabs API error: ${errorJson.detail.message}`;
+      } else if (errorJson.detail) {
+        errorMessage = `ElevenLabs API error: ${JSON.stringify(errorJson.detail)}`;
+      }
+    } catch {
+      if (errorText) {
+        errorMessage = `ElevenLabs API error: ${errorText}`;
+      }
+    }
+
+    throw new Error(errorMessage);
+  }
+
+  log("Speech generated successfully", "success");
+
+  return response.arrayBuffer();
+}
+
+// ============================================================================ 
+// OpenAI TTS API Integration
+// ============================================================================ 
+async function generateWithOpenAI(options: TTSOptions): Promise<ArrayBuffer> {
+  const apiKey = process.env.OPENAI_API_KEY;
+
+  if (!apiKey) {
+    throw new Error("OPENAI_API_KEY environment variable is not set");
+  }
+
+  // Resolve voice
+  const voice = OPENAI_VOICES.includes(options.voice.toLowerCase()) ? options.voice.toLowerCase() : "alloy";
+
+  // Resolve model
+  const modelConfig = OPENAI_MODELS[options.model] || OPENAI_MODELS["tts-1-hd"];
+  const modelId = modelConfig.id;
+
+  // Resolve format - OpenAI supports mp3, opus, aac, flac, wav, pcm
+  let format = options.format.toLowerCase();
+  if (format === "ogg") format = "opus";
+  if (!OPENAI_FORMATS.includes(format)) format = "mp3";
+
+  log(`Provider: OpenAI TTS`);
+  log(`Voice: ${voice}`);
+  log(`Model: ${modelId} - ${modelConfig.description}`);
+  log(`Format: ${format}`);
+  log(`Speed: ${options.speed}`);
+  log(`Text length: ${options.text.length} characters`);
+
+  const response = await fetch("https://api.openai.com/v1/audio/speech", {
+    method: "POST",
+    headers: {
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${apiKey}`,
+    },
+    body: JSON.stringify({
+      model: modelId,
+      input: options.text,
+      voice: voice,
+      response_format: format,
+      speed: options.speed,
+    }),
+  });
+
+  if (!response.ok) {
+    const error = await response.json();
+    throw new Error(`OpenAI API error: ${error.error?.message || response.statusText}`);
+  }
+
+  log("Speech generated successfully", "success");
+
+  return response.arrayBuffer();
+}
+
+// ============================================================================ 
+// Main Generate Function
+// ============================================================================ 
+async function generateSpeech(options: TTSOptions): Promise<ArrayBuffer> {
+  const provider = getProvider(options.model);
+
+  switch (provider) {
+    case "openai":
+      return generateWithOpenAI(options);
+    case "elevenlabs":
+    default:
+      return generateWithElevenLabs(options);
+  }
+}
+
+// ============================================================================ 
+// Main CLI Handler
+// ============================================================================ 
+async function main() {
+  const { values, positionals } = parseArgs({
+    args: Bun.argv.slice(2),
+    options: {
+      voice: { type: "string", default: "rachel" },
+      model: { type: "string", default: "v2" },
+      format: { type: "string", default: "mp3" },
+      quality: { type: "string", default: "high" },
+      stability: { type: "string", default: "0.5" },
+      clarity: { type: "string", default: "0.75" },
+      style: { type: "string", default: "0" },
+      speed: { type: "string", default: "1.0" },
+      language: { type: "string" },
+      file: { type: "string" },
+      output: { type: "string", short: "o" },
+      help: { type: "boolean", short: "h" },
+    },
+    allowPositionals: true,
+  });
+
+  // Show help
+  if (values.help) {
+    console.log(`
+╔═══════════════════════════════════════════════════════════════════════════════╗
+║                      TEXT TO SPEECH - Multi-Provider AI                        ║
+║              Convert text to natural, lifelike speech                          ║
+╚═══════════════════════════════════════════════════════════════════════════════╝
+
+USAGE:
+  skills run generate-audio -- "<text>" [options]
+  skills run generate-audio -- --file ./script.txt [options]
+
+═══════════════════════════════════════════════════════════════════════════════
+                              MODELS
+═══════════════════════════════════════════════════════════════════════════════
+
+  ElevenLabs (--model):
+    v3        Eleven v3 (Alpha) - Most expressive, supports audio tags ⭐
+    v2        Multilingual v2 - Highest quality, 29 languages (default)
+    turbo     Turbo v2.5 - Low latency (~300ms), English optimized
+    flash     Flash v2.5 - Ultra-low latency (75ms), real-time apps
+
+  OpenAI TTS (--model):
+    tts-1-hd  OpenAI TTS HD - Highest quality
+    tts-1     OpenAI TTS - Fast, good quality
+
+═══════════════════════════════════════════════════════════════════════════════
+                              VOICES
+═══════════════════════════════════════════════════════════════════════════════
+
+  ElevenLabs Voices (40+ premium voices):
+    Female: rachel, domi, bella, elli, charlotte, matilda, dorothy, sarah,
+            emily, gigi, freya, grace, lily, serena, nicole, jessie, glinda, mimi
+    Male:   drew, clyde, paul, dave, fin, antoni, thomas, charlie, george,
+            callum, patrick, harry, liam, josh, arnold, matthew, james, joseph,
+            jeremy, michael, ethan, chris, daniel, adam, bill, sam, giovanni
+
+  OpenAI Voices (--voice with --model tts-1 or tts-1-hd):
+    alloy     Neutral, balanced
+    echo      Warm, conversational
+    fable     British, narrative
+    onyx      Deep, authoritative
+    nova      Energetic, bright
+    shimmer   Soft, gentle
+
+═══════════════════════════════════════════════════════════════════════════════
+                        ELEVEN V3 AUDIO TAGS ⭐
+═══════════════════════════════════════════════════════════════════════════════
+
+  V3 supports inline audio tags for expressive speech. Use --model v3
+
+  EMOTIONS:     [excited] [nervous] [frustrated] [calm] [curious] [happily]
+                [angry] [sad] [sorrowful] [crying] [mischievously]
+
+  REACTIONS:    [laughs] [sighs] [gasps] [gulps] [clears throat] [sniffs]
+
+  DELIVERY:     [whispers] [shouts] [softly] [loudly]
+
+  PACING:       [pauses] [hesitates] [stammers]
+
+  TONE:         [cheerfully] [flatly] [deadpan] [playfully] [sarcastically]
+
+  Example:
+    skills run generate-audio -- "[sighs] I can't believe it. [excited] We won!" --model v3
+
+═══════════════════════════════════════════════════════════════════════════════
+
+FORMATS (--format):
+  mp3     Standard MP3 (default)
+  wav     Uncompressed WAV
+  ogg     Ogg Vorbis (ElevenLabs) / Opus (OpenAI)
+  flac    Lossless FLAC
+  aac     AAC (OpenAI only)
+
+QUALITY (--quality for ElevenLabs):
+  low     32 kbps - Small files
+  medium  64 kbps - Balanced
+  high    128 kbps - Good quality (default)
+  ultra   192 kbps - Best quality
+
+VOICE SETTINGS (ElevenLabs only):
+  --stability <0-1>   Voice stability (default: 0.5)
+                      Lower = more expressive, Higher = more consistent
+  --clarity <0-1>     Clarity + similarity boost (default: 0.75)
+  --style <0-1>       Style exaggeration (default: 0)
+  --speed <0.5-2>     Speech speed multiplier (default: 1.0)
+
+OTHER OPTIONS:
+  --language <code>   ISO 639-1 language code (auto-detect by default)
+  --file <path>       Read text from file instead of argument
+  --output, -o <path> Custom output path
+
+EXAMPLES:
+  # ElevenLabs - Default v2 model
+  skills run generate-audio -- "Hello, welcome to our podcast!"
+
+  # ElevenLabs V3 with audio tags
+  skills run generate-audio -- "[whispers] Listen... [excited] This is amazing!" --model v3
+
+  # OpenAI TTS HD with nova voice
+  skills run generate-audio -- "Welcome to the future of AI" --model tts-1-hd --voice nova
+
+  # OpenAI TTS fast generation
+  skills run generate-audio -- "Quick announcement" --model tts-1 --voice onyx
+
+  # ElevenLabs professional narration
+  skills run generate-audio -- "In today's episode..." --voice george --quality ultra
+
+  # ElevenLabs fast turbo for real-time
+  skills run generate-audio -- "Breaking news!" --voice drew --model turbo
+
+  # OpenAI with slower speed
+  skills run generate-audio -- "Take your time..." --model tts-1-hd --voice shimmer --speed 0.8
+");
+    process.exit(0);
+  }
+
+  // Get text from argument or file
+  let text = positionals.join(" ");
+
+  if (values.file) {
+    if (!existsSync(values.file as string)) {
+      log(`File not found: ${values.file}`, "error");
+      process.exit(1);
+    }
+    text = readFileSync(values.file as string, "utf-8").trim();
+    log(`Read ${text.length} characters from ${values.file}`);
+  }
+
+  if (!text) {
+    log("Please provide text to convert or use --file <path>", "error");
+    console.log("Run with --help for usage information");
+    process.exit(1);
+  }
+
+  try {
+    // Determine output path
+    const timestamp = new Date()
+      .toISOString()
+      .replace(/[:.]/g, "_")
+      .replace(/-/g, "_")
+      .slice(0, 19)
+      .toLowerCase();
+    const textSlug = slugify(text);
+    const voiceSlug = values.voice || "rachel";
+    const modelSlug = values.model || "v2";
+    const provider = getProvider(values.model as string);
+
+    // Determine format extension
+    let format = (values.format as string)?.toLowerCase() || "mp3";
+    if (provider === "openai" && format === "ogg") format = "opus";
+    const extension = FORMAT_EXTENSIONS[format] || "mp3";
+
+    const defaultOutput = join(EXPORTS_DIR, `export_${timestamp}_${textSlug}_${voiceSlug}_${modelSlug}.${extension}`);
+    const outputPath = (values.output as string) || defaultOutput;
+
+    log(`Session ID: ${SESSION_ID}`);
+    log(`Output will be saved to: ${outputPath}`);
+
+    // Generate speech
+    const audioBuffer = await generateSpeech({
+      text,
+      provider,
+      voice: values.voice as string,
+      model: values.model as string,
+      format: format,
+      quality: values.quality as string,
+      stability: parseFloat(values.stability as string),
+      clarity: parseFloat(values.clarity as string),
+      style: parseFloat(values.style as string),
+      speed: parseFloat(values.speed as string),
+      language: values.language as string | undefined,
+    });
+
+    // Save audio file
+    ensureDir(dirname(outputPath));
+    writeFileSync(outputPath, Buffer.from(audioBuffer));
+
+    const fileSizeKB = Math.round(audioBuffer.byteLength / 1024);
+    log(`Audio saved: ${outputPath} (${fileSizeKB} KB)`, "success");
+
+    console.log(`
+╔═══════════════════════════════════════════════════════════════════════════════╗
+║                         ✨ SPEECH GENERATED SUCCESSFULLY                       ║
+╠═══════════════════════════════════════════════════════════════════════════════╣
+║  📁 Output: ${outputPath.padEnd(63)}║
+║  📊 Size: ${(fileSizeKB + " KB").padEnd(66)}║
+║  🤖 Provider: ${(PROVIDERS[provider].name).padEnd(61)}║
+║  📋 Log: ${join(LOGS_DIR, `log_${SESSION_TIMESTAMP}_${SESSION_ID}.log`).padEnd(66)}║
+╚═══════════════════════════════════════════════════════════════════════════════╝
+`);
+  } catch (error) {
+    log(`Error: ${error instanceof Error ? error.message : "Unknown error"}`, "error");
+    process.exit(1);
+  }
+}
+
+main();

package/skills/skill-generate-audio/tsconfig.json

@@ -0,0 +1,18 @@
+{
+  "compilerOptions": {
+    "target": "ES2022",
+    "module": "ES2022",
+    "moduleResolution": "bundler",
+    "strict": true,
+    "esModuleInterop": true,
+    "skipLibCheck": true,
+    "forceConsistentCasingInFileNames": true,
+    "resolveJsonModule": true,
+    "declaration": true,
+    "declarationMap": true,
+    "sourceMap": true,
+    "outDir": "./dist",
+    "types": ["bun-types"]
+  },
+  "include": ["src", "bin"]
+}

package/skills/skill-generate-book-cover/CLAUDE.md

@@ -0,0 +1,19 @@
+# skill-generate-book-cover
+
+Generate Book Cover skill.
+
+## Tech Stack
+
+- Runtime: Bun
+- Language: TypeScript
+- CLI: Commander.js
+
+## CLI
+
+```bash
+skill-generate-book-cover run
+```
+
+## Data Directory
+
+~/.skill/skill-generate-book-cover/