mulmocast 0.0.24 → 0.0.25

package/README.md

@@ -103,6 +103,14 @@ GOOGLE_PROJECT_ID=your_google_project_id
 
 See also [pre-requisites for Google's image generation model](./docs/pre-requisites-google.md)
 
+#### (Optional) For AI providers
+```bash
+# For Anthropic Claude (htmlPrompt feature)
+ANTHROPIC_API_TOKEN=your_anthropic_api_token
+```
+
+For htmlPrompt configuration, see [docs/image.md](./docs/image.md#2-htmlprompt).
+
 #### (Optional) For Movie models
 ```bash
 REPLICATE_API_TOKEN=your_replicate_api_key

package/lib/actions/images.js

@@ -77,6 +77,9 @@ export const imagePluginAgent = async (namedInputs) => {
 };
 const htmlImageGeneratorAgent = async (namedInputs) => {
     const { html, file, canvasSize } = namedInputs;
+    // Save HTML file
+    const htmlFile = file.replace(/\.[^/.]+$/, ".html");
+    await fs.promises.writeFile(htmlFile, html, "utf8");
     await renderHTMLToImage(html, file, canvasSize.width, canvasSize.height);
 };
 const beat_graph_data = {
@@ -115,12 +118,13 @@ const beat_graph_data = {
             if: ":preprocessor.htmlPrompt",
             defaultValue: {},
             agent: ":htmlImageAgentInfo.agent",
-            params: {
-                mode: ":htmlImageAgentInfo.model",
-            },
             inputs: {
                 prompt: ":preprocessor.htmlPrompt",
                 system: ":preprocessor.htmlImageSystemPrompt",
+                params: {
+                    model: ":htmlImageAgentInfo.model",
+                    max_tokens: ":htmlImageAgentInfo.max_tokens",
+                },
             },
         },
         htmlImageGenerator: {

package/lib/cli/commands/tool/scripting/builder.d.ts

@@ -16,7 +16,7 @@ export declare const builder: (yargs: Argv) => Argv<{
 } & {
     s: string;
 } & {
-    llm: "anthropic" | "openAI" | "gemini" | "groq" | undefined;
+    llm: "openai" | "anthropic" | "gemini" | "groq" | undefined;
 } & {
     llm_model: string | undefined;
 }>;

package/lib/cli/commands/tool/story_to_script/builder.d.ts

@@ -10,7 +10,7 @@ export declare const builder: (yargs: Argv) => Argv<{
 } & {
     beats_per_scene: number;
 } & {
-    llm: "anthropic" | "openAI" | "gemini" | "groq" | undefined;
+    llm: "openai" | "anthropic" | "gemini" | "groq" | undefined;
 } & {
     llm_model: string | undefined;
 } & {

package/lib/methods/mulmo_presentation_style.js

@@ -1,5 +1,5 @@
 import "dotenv/config";
-import { userAssert } from "../utils/utils.js";
+import { userAssert, llmConfig } from "../utils/utils.js";
 import { text2ImageProviderSchema, text2HtmlImageProviderSchema, text2SpeechProviderSchema, mulmoCanvasDimensionSchema } from "../types/schema.js";
 import { defaultOpenAIImageModel } from "../utils/const.js";
 const defaultTextSlideStyles = [
@@ -72,16 +72,13 @@ export const MulmoPresentationStyleMethods = {
     },
     getHtmlImageAgentInfo(presentationStyle) {
         const provider = text2HtmlImageProviderSchema.parse(presentationStyle.htmlImageParams?.provider);
-        const agent = provider === "anthropic" ? "anthropicAgent" : "openAIAgent";
-        const model = presentationStyle.htmlImageParams?.model
-            ? presentationStyle.htmlImageParams?.model
-            : provider === "anthropic"
-                ? "claude-3-7-sonnet-20250219"
-                : "gpt-4o-mini";
+        const defaultConfig = llmConfig[provider];
+        const model = presentationStyle.htmlImageParams?.model ? presentationStyle.htmlImageParams?.model : defaultConfig.defaultModel;
         return {
             provider,
-            agent,
+            agent: defaultConfig.agent,
             model,
+            max_tokens: defaultConfig.max_tokens,
         };
     },
     getImageType(_, beat) {

package/lib/types/type.d.ts

@@ -83,6 +83,7 @@ export type Text2HtmlAgentInfo = {
     provider: Text2HtmlImageProvider;
     agent: string;
     model: string;
+    max_tokens: number;
 };
 export type BeatMediaType = "movie" | "image";
 export type StoryToScriptGenerateMode = (typeof storyToScriptGenerateMode)[keyof typeof storyToScriptGenerateMode];

package/lib/utils/prompt.js

@@ -128,7 +128,7 @@ export const htmlImageSystemPrompt = (canvasSize) => {
     return [
         "Based on the provided information, create a single slide HTML page using Tailwind CSS.",
         `The view port size is ${canvasSize.width}x${canvasSize.height}. Make sure the HTML fits within the view port.`,
-        "If charts are needed, use Chart.js to present them in a clean and visually appealing way.",
+        "If charts are needed, use Chart.js to present them in a clean and visually appealing way (with animation:false to disable animation).",
         "Include a balanced mix of comments, graphs, and illustrations to enhance visual impact.",
         "Output only the HTML code. Do not include any comments, explanations, or additional information outside the HTML.",
         "If data is provided, use it effectively to populate the slide.",

package/lib/utils/utils.d.ts

@@ -1,6 +1,6 @@
 import { MulmoBeat, MulmoStudioMultiLingualData } from "../types/index.js";
 import type { ConfigDataDictionary, DefaultConfigData } from "graphai";
-export declare const llm: readonly ["openAI", "anthropic", "gemini", "groq"];
+export declare const llm: readonly ["openai", "anthropic", "gemini", "groq"];
 export type LLM = (typeof llm)[number];
 export declare const llmConfig: Record<LLM, {
     agent: string;

package/lib/utils/utils.js

@@ -1,7 +1,7 @@
 import * as crypto from "crypto";
-export const llm = ["openAI", "anthropic", "gemini", "groq"];
+export const llm = ["openai", "anthropic", "gemini", "groq"];
 export const llmConfig = {
-    openAI: {
+    openai: {
         agent: "openAIAgent",
         defaultModel: "gpt-4o",
         max_tokens: 8192,
@@ -23,10 +23,10 @@ export const llmConfig = {
     },
 };
 export const llmPair = (_llm, _model) => {
-    const llmKey = _llm ?? "openAI";
-    const agent = llmConfig[llmKey]?.agent ?? llmConfig.openAI.agent;
-    const model = _model ?? llmConfig[llmKey]?.defaultModel ?? llmConfig.openAI.defaultModel;
-    const max_tokens = llmConfig[llmKey]?.max_tokens ?? llmConfig.openAI.max_tokens;
+    const llmKey = _llm ?? "openai";
+    const agent = llmConfig[llmKey]?.agent ?? llmConfig.openai.agent;
+    const model = _model ?? llmConfig[llmKey]?.defaultModel ?? llmConfig.openai.defaultModel;
+    const max_tokens = llmConfig[llmKey]?.max_tokens ?? llmConfig.openai.max_tokens;
     return { agent, model, max_tokens };
 };
 export const chunkArray = (array, size = 3) => {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mulmocast",
-  "version": "0.0.24",
+  "version": "0.0.25",
   "description": "",
   "type": "module",
   "main": "lib/index.js",
@@ -68,22 +68,22 @@
     "@graphai/stream_agent_filter": "^2.0.2",
     "@graphai/vanilla": "^2.0.4",
     "@graphai/vanilla_node_agents": "^2.0.1",
-    "@tavily/core": "^0.5.7",
-    "canvas": "^3.1.0",
+    "@tavily/core": "^0.5.8",
+    "canvas": "^3.1.2",
     "clipboardy": "^4.0.0",
-    "dotenv": "^16.4.7",
+    "dotenv": "^17.0.0",
     "fluent-ffmpeg": "^2.1.3",
     "google-auth-library": "^9.15.1",
     "graphai": "^2.0.9",
-    "inquirer": "^12.6.1",
+    "inquirer": "^12.6.3",
     "marked": "^15.0.12",
     "ora": "^8.2.0",
-    "puppeteer": "^24.10.2",
+    "puppeteer": "^24.11.0",
     "replicate": "^1.0.1",
     "yaml": "^2.8.0",
     "yargs": "^17.7.2",
     "zod": "^3.25.67",
-    "zod-to-json-schema": "^3.24.5"
+    "zod-to-json-schema": "^3.24.6"
   },
   "devDependencies": {
     "@anatine/zod-mock": "^3.14.0",
@@ -91,14 +91,14 @@
     "@receptron/test_utils": "^2.0.0",
     "@types/fluent-ffmpeg": "^2.1.26",
     "@types/yargs": "^17.0.33",
-    "eslint": "^9.29.0",
+    "eslint": "^9.30.0",
     "eslint-config-prettier": "^10.1.5",
-    "eslint-plugin-prettier": "^5.5.0",
-    "prettier": "^3.3.3",
+    "eslint-plugin-prettier": "^5.5.1",
+    "prettier": "^3.6.2",
     "ts-node": "^10.9.2",
     "tsx": "^4.20.3",
     "typescript": "^5.7.3",
-    "typescript-eslint": "^8.34.1"
+    "typescript-eslint": "^8.35.0"
   },
   "engines": {
     "node": ">=18.0.0"