mulmocast 2.1.9 → 2.1.11

package/lib/actions/bundle.js

@@ -127,7 +127,7 @@ export const mulmoViewerBundle = async (context) => {
     });
     // BGM
     const bgmFileName = await processBgm(context.studio?.script.audioParams?.bgm, dir, zipper);
-    const bundleData = { beats: resultJson, bgmSource: bgmFileName };
+    const bundleData = { beats: resultJson, bgmSource: bgmFileName, title: context.studio.script.title };
     fs.writeFileSync(path.resolve(dir, viewJsonFileName), JSON.stringify(bundleData, null, 2));
     zipper.addFile(path.resolve(dir, viewJsonFileName));
     if (isZip) {

package/lib/agents/tts_gemini_agent.js

@@ -3,9 +3,16 @@ import { GoogleGenAI } from "@google/genai";
 import { provider2TTSAgent } from "../utils/provider2agent.js";
 import { agentIncorrectAPIKeyError, apiKeyMissingError, agentGenerationError, audioAction, audioFileTarget, getGenAIErrorReason, } from "../utils/error_cause.js";
 import { pcmToMp3 } from "../utils/ffmpeg_utils.js";
+const getPrompt = (text, instructions) => {
+    // https://ai.google.dev/gemini-api/docs/speech-generation?hl=ja#controllable
+    if (instructions) {
+        return `### DIRECTOR'S NOTES\n${instructions}\n\n#### TRANSCRIPT\n${text}`;
+    }
+    return text;
+};
 export const ttsGeminiAgent = async ({ namedInputs, params, config, }) => {
     const { text } = namedInputs;
-    const { model, voice, suppressError } = params;
+    const { model, voice, suppressError, instructions } = params;
     const apiKey = config?.apiKey;
     if (!apiKey) {
         throw new Error("Google GenAI API key is required (GEMINI_API_KEY)", {
@@ -16,7 +23,7 @@ export const ttsGeminiAgent = async ({ namedInputs, params, config, }) => {
         const ai = new GoogleGenAI({ apiKey });
         const response = await ai.models.generateContent({
             model: model ?? provider2TTSAgent.gemini.defaultModel,
-            contents: [{ parts: [{ text }] }],
+            contents: [{ parts: [{ text: getPrompt(text, instructions) }] }],
             config: {
                 responseModalities: ["AUDIO"],
                 speechConfig: {

package/lib/types/agent.d.ts

@@ -122,6 +122,7 @@ export type KotodamaTTSAgentParams = TTSAgentParams & {
 export type GoogleTTSAgentParams = TTSAgentParams & {
     speed: number;
     model: string;
+    instructions: string;
 };
 export type ElevenlabsTTSAgentParams = TTSAgentParams & {
     model: string;

package/lib/types/type.d.ts

@@ -154,4 +154,5 @@ export type MulmoViewerBeat = {
 export type MulmoViewerData = {
     beats: MulmoViewerBeat[];
     bgmSource?: string;
+    title?: string;
 };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mulmocast",
-  "version": "2.1.9",
+  "version": "2.1.11",
   "description": "",
   "type": "module",
   "main": "lib/index.node.js",
@@ -84,13 +84,13 @@
     "@graphai/stream_agent_filter": "^2.0.2",
     "@graphai/vanilla": "^2.0.12",
     "@graphai/vanilla_node_agents": "^2.0.4",
-    "@inquirer/input": "^4.3.0",
-    "@inquirer/select": "^4.4.1",
+    "@inquirer/input": "^5.0.3",
+    "@inquirer/select": "^5.0.3",
     "@modelcontextprotocol/sdk": "^1.25.1",
     "@mozilla/readability": "^0.6.0",
     "@tavily/core": "^0.5.11",
     "archiver": "^7.0.1",
-    "clipboardy": "^4.0.0",
+    "clipboardy": "^5.0.2",
     "dotenv": "^17.2.3",
     "fluent-ffmpeg": "^2.1.3",
     "graphai": "^2.0.16",
@@ -98,7 +98,7 @@
     "marked": "^17.0.1",
     "mulmocast-vision": "^1.0.8",
     "ora": "^9.0.0",
-    "puppeteer": "^24.33.0",
+    "puppeteer": "^24.34.0",
     "replicate": "^1.4.0",
     "yaml": "^2.8.2",
     "yargs": "^18.0.0",
@@ -117,7 +117,7 @@
     "prettier": "^3.7.4",
     "tsx": "^4.21.0",
     "typescript": "^5.9.3",
-    "typescript-eslint": "^8.50.0"
+    "typescript-eslint": "^8.50.1"
   },
   "engines": {
     "node": ">=20.0.0"

package/scripts/test/test_audio_gemini.json

@@ -14,6 +14,14 @@
         "provider": "gemini",
         "model": "gemini-2.5-pro-preview-tts",
         "voiceId": "Puck"
+      },
+      "Presenter2": {
+        "provider": "gemini",
+        "model": "gemini-2.5-pro-preview-tts",
+        "voiceId": "leda",
+        "speechOptions": {
+          "instruction": "Role: You are a classic Tsundere character. Tone: Sharp, impatient, and defensive, but with underlying hesitation that suggests you actually care deeply. Voice Quality: Higher pitch, energetic, and slightly haughty."
+        }
       }
     }
   },
@@ -52,7 +60,7 @@
       }
     },
     {
-      "speaker": "Presenter",
+      "speaker": "Presenter2",
       "text": "Hello, I'm a presenter. I have a whisper instruction.",
       "speechOptions": {
         "instruction": "Whisper softly, like a pillow talk."

package/scripts/test/test_media.json

@@ -60,7 +60,7 @@
       "speaker": "Presenter",
       "text": "",
       "duration": 0.5,
-      "id": "{1A57B3F5-B6CB-4948-96BB-6F018DCCBBD4}",
+      "id": "1A57B3F5-B6CB-4948-96BB-6F018DCCBBD4",
       "image": {
         "type": "textSlide",
         "slide": {