TTS

import { defineModels } from "@scrydon/sdk-authoring/extensions/define";
import type { TTSCapabilityConfig } from "@scrydon/sdk-authoring/extensions/capabilities/tts";

const ttsCapability: TTSCapabilityConfig = {
  models: [
    { id: "tts-standard", name: "Standard TTS", voices: 6, benchmarks: [{ name: "MOS", score: 3.5, source: "Internal" }] },
    { id: "tts-hd",       name: "HD TTS",       voices: 6, benchmarks: [{ name: "MOS", score: 4.2, source: "Internal" }] },
  ],
  defaultModel: "tts-standard",

  // Batch mode — send text, get audio back
  async synthesize(request) {
    const response = await fetch("https://api.example.com/v1/synthesize", {
      method: "POST",
      headers: {
        Authorization: `Bearer ${request.apiKey}`,
        "Content-Type": "application/json",
      },
      body: JSON.stringify({
        text: request.text,
        voice: request.voice,
        model: request.model,
        speed: request.speed,
      }),
    });
    return {
      audioBuffer: await response.arrayBuffer(),
      format: "mp3",
      mimeType: "audio/mpeg",
    };
  },

  // Realtime streaming (optional)
  realtime: {
    protocol: "websocket",
    async createSession(config) {
      // The wire protocol is your vendor's; return a RealtimeSession.
      const socket = new WebSocket("wss://api.example.com/v1/realtime-tts");
      await new Promise((resolve, reject) => {
        socket.addEventListener("open", resolve, { once: true });
        socket.addEventListener("error", reject, { once: true });
      });
      socket.send(JSON.stringify({ type: "start", apiKey: config.apiKey, voice: config.voice, model: config.model }));
      return {
        sessionId: crypto.randomUUID(),
        send: (chunk) => socket.send(chunk),
        onMessage: (handler) =>
          socket.addEventListener("message", (event) =>
            handler({ type: "audio", data: event.data, timestamp: Date.now() })
          ),
        close: async () => socket.close(),
      };
    },
    features: { streamingInput: true, streamingOutput: true },
  },
};

// One model family per runtime; list it under defineExtension()'s provides.models.
export const ttsModels = defineModels({
  id: "my-service-tts",
  name: "My Service Text-to-Speech",
  family: "tts",
  runtime: ttsCapability,
});