feat: add local audio test mode

src/services/elevenlabs-stt.ts

@@ -1,6 +1,6 @@
 import WebSocket from "ws";
 
-import type { AppConfig } from "../config.js";
+import type { AssistantRuntimeConfig } from "../config.js";
 
 interface ElevenLabsMessage {
   message_type?: string;
@@ -13,7 +13,7 @@ const NON_FATAL_ERROR_TYPES = new Set([
 ]);
 
 export class ElevenLabsSttService {
-  constructor(private readonly config: AppConfig) {}
+  constructor(private readonly config: AssistantRuntimeConfig) {}
 
   async transcribePcm16(pcm16MonoAudio: Buffer): Promise<string | null> {
     if (pcm16MonoAudio.byteLength === 0) {

src/services/elevenlabs-tts.ts

@@ -2,24 +2,23 @@ import { Readable } from "node:stream";
 
 import ffmpegStatic from "ffmpeg-static";
 import prism from "prism-media";
-import { StreamType, createAudioResource, type AudioResource } from "@discordjs/voice";
 
-import type { AppConfig } from "../config.js";
+import type { AssistantRuntimeConfig } from "../config.js";
 
-export interface PreparedSpeechPlayback {
-  resource: AudioResource;
+export interface PreparedSpeechAudio {
+  stream: Readable;
   dispose: () => void;
 }
 
 export class ElevenLabsTtsService {
-  constructor(private readonly config: AppConfig) {
+  constructor(private readonly config: AssistantRuntimeConfig) {
     const resolvedFfmpegPath = ffmpegStatic as unknown as string | null;
     if (resolvedFfmpegPath && !process.env.FFMPEG_PATH) {
       process.env.FFMPEG_PATH = resolvedFfmpegPath;
     }
   }
 
-  async preparePlayback(text: string, signal?: AbortSignal): Promise<PreparedSpeechPlayback> {
+  async preparePlayback(text: string, signal?: AbortSignal): Promise<PreparedSpeechAudio> {
     const url = new URL(`https://api.elevenlabs.io/v1/text-to-speech/${this.config.ELEVENLABS_VOICE_ID}/stream`);
     url.searchParams.set("output_format", "mp3_44100_128");
     url.searchParams.set("enable_logging", "false");
@@ -68,12 +67,8 @@ export class ElevenLabsTtsService {
 
     input.pipe(ffmpeg);
 
-    const resource = createAudioResource(ffmpeg, {
-      inputType: StreamType.Raw,
-    });
-
     return {
-      resource,
+      stream: ffmpeg,
       dispose: () => {
         input.destroy();
         ffmpeg.destroy();

src/services/openai-llm.ts

@@ -1,6 +1,6 @@
 import OpenAI from "openai";
 
-import type { AppConfig } from "../config.js";
+import type { AssistantRuntimeConfig } from "../config.js";
 import type { ConversationMemory, UserUtterance } from "./conversation.js";
 
 const ASSISTANT_INSTRUCTIONS = [
@@ -30,7 +30,7 @@ function normalizeReply(text: string): string {
 export class OpenAiLlmService {
   private readonly client: OpenAI;
 
-  constructor(private readonly config: AppConfig) {
+  constructor(private readonly config: AssistantRuntimeConfig) {
     this.client = new OpenAI({
       apiKey: this.config.OPENAI_API_KEY,
     });