mulmocast 1.2.28 → 1.2.30

package/lib/actions/html.d.ts

@@ -0,0 +1,3 @@
+import { MulmoStudioContext } from "../types/index.js";
+export declare const htmlFilePath: (context: MulmoStudioContext) => string;
+export declare const html: (context: MulmoStudioContext, imageWidth?: string) => Promise<void>;

package/lib/actions/html.js

@@ -0,0 +1,62 @@
+import fs from "fs";
+import path from "path";
+import { isNull } from "graphai";
+import { localizedText } from "../utils/utils.js";
+import { writingMessage } from "../utils/file.js";
+import { MulmoStudioContextMethods } from "../methods/mulmo_studio_context.js";
+const generateHtmlContent = (context, imageWidth) => {
+    const { studio, multiLingual, lang = "en" } = context;
+    const title = studio.script.title || "MulmoCast Content";
+    const description = studio.script.description || "";
+    let html = `<h1>${title}</h1>\n\n`;
+    if (description) {
+        html += `${description}\n\n`;
+    }
+    studio.script.beats.forEach((beat, index) => {
+        const text = localizedText(beat, multiLingual?.[index], lang);
+        const studioBeat = studio.beats[index];
+        if (text.trim() || studioBeat?.html || studioBeat?.imageFile) {
+            if (studioBeat?.html) {
+                html += `${studioBeat.html}\n\n`;
+            }
+            else if (studioBeat?.imageFile && isNull(studioBeat.html)) {
+                const imagePath = path.relative(context.fileDirs.outDirPath, studioBeat.imageFile);
+                const altText = `Beat ${index + 1}`;
+                if (imageWidth) {
+                    // Use HTML img tag for width control
+                    html += `<img src="${imagePath}" alt="${altText}" width="${imageWidth}" />\n\n`;
+                }
+                else {
+                    // Use standard html image syntax
+                    html += `<img src="${imagePath}" alt="${altText}" />\n\n`;
+                }
+            }
+            if (text.trim()) {
+                html += `${text}\n\n`;
+            }
+        }
+    });
+    return html;
+};
+export const htmlFilePath = (context) => {
+    const { studio, fileDirs, lang = "en" } = context;
+    // Add language suffix only when target language is different from script's original language
+    const langSuffix = studio.script.lang !== lang ? `_${lang}` : "";
+    const filename = `${studio.filename}${langSuffix}.html`;
+    return path.join(fileDirs.outDirPath, filename);
+};
+const generateHtml = async (context, imageWidth) => {
+    const outputHtmlPath = htmlFilePath(context);
+    const htmlContent = generateHtmlContent(context, imageWidth);
+    fs.writeFileSync(outputHtmlPath, htmlContent, "utf8");
+    writingMessage(outputHtmlPath);
+};
+export const html = async (context, imageWidth) => {
+    try {
+        MulmoStudioContextMethods.setSessionState(context, "html", true);
+        await generateHtml(context, imageWidth);
+    }
+    finally {
+        MulmoStudioContextMethods.setSessionState(context, "html", false);
+    }
+};

package/lib/actions/image_agents.d.ts

@@ -12,9 +12,9 @@ export declare const imagePreprocessAgent: (namedInputs: {
     htmlImageSystemPrompt: string;
 } | {
     imagePath: string | undefined;
+    movieFile: string | undefined;
     referenceImageForMovie: string | undefined;
     imageParams: MulmoImageParams;
-    movieFile: string | undefined;
     soundEffectFile?: string;
     soundEffectPrompt?: string;
     soundEffectModel?: string;
@@ -35,6 +35,7 @@ export declare const imagePreprocessAgent: (namedInputs: {
         agent: string;
         movieParams: MulmoMovieParams;
     };
+    markdown?: string;
     htmlPrompt?: undefined;
     htmlImageFile?: undefined;
     htmlPath?: undefined;
@@ -64,6 +65,7 @@ export declare const imagePreprocessAgent: (namedInputs: {
         agent: string;
         movieParams: MulmoMovieParams;
     };
+    markdown?: string;
     htmlPrompt?: undefined;
     htmlImageFile?: undefined;
     htmlPath?: undefined;
@@ -96,6 +98,7 @@ export declare const imagePreprocessAgent: (namedInputs: {
         agent: string;
         movieParams: MulmoMovieParams;
     };
+    markdown?: string;
     htmlPrompt?: undefined;
     htmlImageFile?: undefined;
     htmlPath?: undefined;

package/lib/actions/image_agents.js

@@ -60,8 +60,17 @@ export const imagePreprocessAgent = async (namedInputs) => {
     if (beat.image) {
         const plugin = MulmoBeatMethods.getPlugin(beat);
         const pluginPath = plugin.path({ beat, context, imagePath, ...htmlStyle(context, beat) });
+        if (plugin.markdown) {
+            returnValue.markdown = plugin.markdown({ beat, context, imagePath, ...htmlStyle(context, beat) });
+        }
+        const isTypeMovie = beat.image.type === "movie";
         // undefined prompt indicates that image generation is not needed
-        return { ...returnValue, imagePath: pluginPath, referenceImageForMovie: pluginPath };
+        return {
+            ...returnValue,
+            imagePath: isTypeMovie ? undefined : pluginPath,
+            movieFile: isTypeMovie ? pluginPath : undefined,
+            referenceImageForMovie: pluginPath,
+        };
     }
     if (beat.moviePrompt && !beat.imagePrompt) {
         return { ...returnValue, imagePath, imageFromMovie: true }; // no image prompt, only movie prompt

package/lib/actions/images.d.ts

@@ -35,9 +35,9 @@ export declare const beat_graph_data: {
                 htmlImageSystemPrompt: string;
             } | {
                 imagePath: string | undefined;
+                movieFile: string | undefined;
                 referenceImageForMovie: string | undefined;
                 imageParams: MulmoImageParams;
-                movieFile: string | undefined;
                 soundEffectFile?: string;
                 soundEffectPrompt?: string;
                 soundEffectModel?: string;
@@ -58,6 +58,7 @@ export declare const beat_graph_data: {
                     agent: string;
                     movieParams: import("../types/type.js").MulmoMovieParams;
                 };
+                markdown?: string;
                 htmlPrompt?: undefined;
                 htmlImageFile?: undefined;
                 htmlPath?: undefined;
@@ -87,6 +88,7 @@ export declare const beat_graph_data: {
                     agent: string;
                     movieParams: import("../types/type.js").MulmoMovieParams;
                 };
+                markdown?: string;
                 htmlPrompt?: undefined;
                 htmlImageFile?: undefined;
                 htmlPath?: undefined;
@@ -119,6 +121,7 @@ export declare const beat_graph_data: {
                     agent: string;
                     movieParams: import("../types/type.js").MulmoMovieParams;
                 };
+                markdown?: string;
                 htmlPrompt?: undefined;
                 htmlImageFile?: undefined;
                 htmlPath?: undefined;
@@ -359,6 +362,7 @@ export declare const beat_graph_data: {
                 lipSyncFile: string;
                 hasMovieAudio: string;
                 htmlImageFile: string;
+                markdown: string;
             };
             output: {
                 imageFile: string;
@@ -367,6 +371,7 @@ export declare const beat_graph_data: {
                 lipSyncFile: string;
                 hasMovieAudio: string;
                 htmlImageFile: string;
+                markdown: string;
             };
             isResult: boolean;
         };

package/lib/actions/images.js

@@ -285,6 +285,7 @@ export const beat_graph_data = {
                 lipSyncFile: ":preprocessor.lipSyncFile",
                 hasMovieAudio: ":audioChecker.hasMovieAudio",
                 htmlImageFile: ":preprocessor.htmlImageFile",
+                markdown: ":preprocessor.markdown",
             },
             output: {
                 imageFile: ".imageFile",
@@ -293,6 +294,7 @@ export const beat_graph_data = {
                 lipSyncFile: ".lipSyncFile",
                 hasMovieAudio: ".hasMovieAudio",
                 htmlImageFile: ".htmlImageFile",
+                markdown: ".markdown",
             },
             isResult: true,
         },

package/lib/actions/index.d.ts

@@ -6,3 +6,5 @@ export * from "./image_agents.js";
 export * from "./movie.js";
 export * from "./pdf.js";
 export * from "./translate.js";
+export * from "./markdown.js";
+export * from "./html.js";

package/lib/actions/index.js

@@ -6,3 +6,5 @@ export * from "./image_agents.js";
 export * from "./movie.js";
 export * from "./pdf.js";
 export * from "./translate.js";
+export * from "./markdown.js";
+export * from "./html.js";

package/lib/actions/markdown.d.ts

@@ -0,0 +1,3 @@
+import { MulmoStudioContext } from "../types/index.js";
+export declare const markdownFilePath: (context: MulmoStudioContext) => string;
+export declare const markdown: (context: MulmoStudioContext, imageWidth?: string) => Promise<void>;

package/lib/actions/markdown.js

@@ -0,0 +1,61 @@
+import fs from "fs";
+import { localizedText } from "../utils/utils.js";
+import { writingMessage } from "../utils/file.js";
+import { MulmoStudioContextMethods } from "../methods/mulmo_studio_context.js";
+import path from "path";
+const generateMarkdownContent = (context, imageWidth) => {
+    const { studio, multiLingual, lang = "en" } = context;
+    const title = studio.script.title || "MulmoCast Content";
+    const description = studio.script.description || "";
+    let markdown = `# ${title}\n\n`;
+    if (description) {
+        markdown += `${description}\n\n`;
+    }
+    studio.script.beats.forEach((beat, index) => {
+        const text = localizedText(beat, multiLingual?.[index], lang);
+        const studioBeat = studio.beats[index];
+        if (text.trim() || studioBeat?.markdown || studioBeat?.imageFile) {
+            if (studioBeat?.markdown) {
+                markdown += `${studioBeat.markdown}\n\n`;
+            }
+            else if (studioBeat?.imageFile) {
+                const imagePath = path.relative(context.fileDirs.outDirPath, studioBeat.imageFile);
+                if (imageWidth) {
+                    // Use HTML img tag for width control
+                    const altText = `Beat ${index + 1}`;
+                    markdown += `<img src="${imagePath}" alt="${altText}" width="${imageWidth}" />\n\n`;
+                }
+                else {
+                    // Use standard markdown image syntax
+                    markdown += `![Beat ${index + 1}](${imagePath})\n\n`;
+                }
+            }
+            if (text.trim()) {
+                markdown += `${text}\n\n`;
+            }
+        }
+    });
+    return markdown;
+};
+export const markdownFilePath = (context) => {
+    const { studio, fileDirs, lang = "en" } = context;
+    // Add language suffix only when target language is different from script's original language
+    const langSuffix = studio.script.lang !== lang ? `_${lang}` : "";
+    const filename = `${studio.filename}${langSuffix}.md`;
+    return path.join(fileDirs.outDirPath, filename);
+};
+const generateMarkdown = async (context, imageWidth) => {
+    const outputMarkdownPath = markdownFilePath(context);
+    const markdownContent = generateMarkdownContent(context, imageWidth);
+    fs.writeFileSync(outputMarkdownPath, markdownContent, "utf8");
+    writingMessage(outputMarkdownPath);
+};
+export const markdown = async (context, imageWidth) => {
+    try {
+        MulmoStudioContextMethods.setSessionState(context, "markdown", true);
+        await generateMarkdown(context, imageWidth);
+    }
+    finally {
+        MulmoStudioContextMethods.setSessionState(context, "markdown", false);
+    }
+};

package/lib/agents/image_replicate_agent.js

@@ -1,9 +1,10 @@
+import { readFileSync } from "fs";
 import { GraphAILogger } from "graphai";
 import Replicate from "replicate";
 import { getAspectRatio } from "./movie_replicate_agent.js";
 import { provider2ImageAgent } from "../utils/provider2agent.js";
 export const imageReplicateAgent = async ({ namedInputs, params, config, }) => {
-    const { prompt } = namedInputs;
+    const { prompt, referenceImages } = namedInputs;
     const { canvasSize } = params;
     const model = params.model ?? provider2ImageAgent.replicate.defaultModel;
     const apiKey = config?.apiKey;
@@ -15,40 +16,14 @@ export const imageReplicateAgent = async ({ namedInputs, params, config, }) => {
     });
     const input = {
         prompt,
-        width: canvasSize.width,
-        height: canvasSize.height,
+        aspect_ratio: getAspectRatio(canvasSize),
     };
-    if (model === "bytedance/seedream-4") {
-        input.size = "custom";
-        if (input.width < 1024) {
-            const ratio = 1024 / input.width;
-            input.width = 1024;
-            input.height = Math.round(input.height * ratio);
-        }
-        if (input.height < 1024) {
-            const ratio = 1024 / input.height;
-            input.width = Math.round(input.width * ratio);
-            input.height = 1024;
-        }
-    }
-    else if (model === "qwen/qwen-image") {
-        input.aspect_ratio = getAspectRatio(canvasSize);
-    }
-    // Add image if provided (for image-to-image generation)
-    /*
-    if (imagePath) {
-      const buffer = readFileSync(imagePath);
-      const base64Image = `data:image/png;base64,${buffer.toString("base64")}`;
-      const start_image = provider2MovieAgent.replicate.modelParams[model]?.start_image;
-      if (start_image === "first_frame_image" || start_image === "image" || start_image === "start_image") {
-        input[start_image] = base64Image;
-      } else if (start_image === undefined) {
-        throw new Error(`Model ${model} does not support image-to-video generation`);
-      } else {
-        input.image = base64Image;
-      }
+    if (referenceImages && referenceImages.length > 0) {
+        input.image_input = referenceImages.map((image) => {
+            const buffer = readFileSync(image);
+            return `data:image/png;base64,${buffer.toString("base64")}`;
+        });
     }
-    */
     try {
         const output = await replicate.run(model, { input });
         // Download the generated video

package/lib/agents/movie_replicate_agent.js

@@ -54,15 +54,17 @@ async function generateMovie(model, apiKey, prompt, imagePath, aspectRatio, dura
     }
 }
 export const getAspectRatio = (canvasSize) => {
-    if (canvasSize.width > canvasSize.height) {
+    const ratio = canvasSize.width / canvasSize.height;
+    const tolerance = 0.1;
+    if (ratio > 4 / 3 + tolerance)
         return "16:9";
-    }
-    else if (canvasSize.width < canvasSize.height) {
-        return "9:16";
-    }
-    else {
+    if (ratio > 4 / 3 - tolerance)
+        return "4:3";
+    if (ratio > 3 / 4 + tolerance)
         return "1:1";
-    }
+    if (ratio > 3 / 4 - tolerance)
+        return "3:4";
+    return "9:16";
 };
 export const movieReplicateAgent = async ({ namedInputs, params, config, }) => {
     const { prompt, imagePath } = namedInputs;

package/lib/cli/bin.js

@@ -10,6 +10,8 @@ import * as audioCmd from "./commands/audio/index.js";
 import * as imagesCmd from "./commands/image/index.js";
 import * as movieCmd from "./commands/movie/index.js";
 import * as pdfCmd from "./commands/pdf/index.js";
+import * as markdownCmd from "./commands/markdown/index.js";
+import * as htmlCmd from "./commands/html/index.js";
 import * as toolCmd from "./commands/tool/index.js";
 import { GraphAILogger } from "graphai";
 const __filename = fileURLToPath(import.meta.url);
@@ -32,6 +34,8 @@ export const main = async () => {
         .command(imagesCmd)
         .command(movieCmd)
         .command(pdfCmd)
+        .command(markdownCmd)
+        .command(htmlCmd)
         .command(toolCmd)
         .demandCommand()
         .strict()

package/lib/cli/commands/html/builder.d.ts

@@ -0,0 +1,16 @@
+import type { Argv } from "yargs";
+export declare const builder: (yargs: Argv) => Argv<{
+    o: string | undefined;
+} & {
+    b: string | undefined;
+} & {
+    l: string | undefined;
+} & {
+    f: boolean;
+} & {
+    p: string | undefined;
+} & {
+    file: string | undefined;
+} & {
+    image_width: string | undefined;
+}>;

package/lib/cli/commands/html/builder.js

@@ -0,0 +1,5 @@
+import { commonOptions } from "../../common.js";
+export const builder = (yargs) => commonOptions(yargs).option("image_width", {
+    describe: "Image width (e.g., 400px, 50%, auto)",
+    type: "string",
+});

package/lib/cli/commands/html/handler.d.ts

@@ -0,0 +1,4 @@
+import { CliArgs } from "../../../types/cli_types.js";
+export declare const handler: (argv: CliArgs<{
+    image_width?: string;
+}>) => Promise<void>;

package/lib/cli/commands/html/handler.js

@@ -0,0 +1,11 @@
+import { images, html } from "../../../actions/index.js";
+import { initializeContext, runTranslateIfNeeded } from "../../helpers.js";
+export const handler = async (argv) => {
+    const context = await initializeContext(argv);
+    if (!context) {
+        process.exit(1);
+    }
+    await runTranslateIfNeeded(context);
+    await images(context);
+    await html(context, argv.image_width);
+};

package/lib/cli/commands/html/index.d.ts

@@ -0,0 +1,4 @@
+export declare const command = "html <file>";
+export declare const desc = "Generate html files";
+export { builder } from "./builder.js";
+export { handler } from "./handler.js";

package/lib/cli/commands/html/index.js

@@ -0,0 +1,4 @@
+export const command = "html <file>";
+export const desc = "Generate html files";
+export { builder } from "./builder.js";
+export { handler } from "./handler.js";

package/lib/cli/commands/markdown/builder.d.ts

@@ -0,0 +1,16 @@
+import type { Argv } from "yargs";
+export declare const builder: (yargs: Argv) => Argv<{
+    o: string | undefined;
+} & {
+    b: string | undefined;
+} & {
+    l: string | undefined;
+} & {
+    f: boolean;
+} & {
+    p: string | undefined;
+} & {
+    file: string | undefined;
+} & {
+    image_width: string | undefined;
+}>;

package/lib/cli/commands/markdown/builder.js

@@ -0,0 +1,5 @@
+import { commonOptions } from "../../common.js";
+export const builder = (yargs) => commonOptions(yargs).option("image_width", {
+    describe: "Image width (e.g., 400px, 50%, auto)",
+    type: "string",
+});

package/lib/cli/commands/markdown/handler.d.ts

@@ -0,0 +1,4 @@
+import { CliArgs } from "../../../types/cli_types.js";
+export declare const handler: (argv: CliArgs<{
+    image_width?: string;
+}>) => Promise<void>;

package/lib/cli/commands/markdown/handler.js

@@ -0,0 +1,11 @@
+import { images, markdown } from "../../../actions/index.js";
+import { initializeContext, runTranslateIfNeeded } from "../../helpers.js";
+export const handler = async (argv) => {
+    const context = await initializeContext(argv);
+    if (!context) {
+        process.exit(1);
+    }
+    await runTranslateIfNeeded(context);
+    await images(context);
+    await markdown(context, argv.image_width);
+};

package/lib/cli/commands/markdown/index.d.ts

@@ -0,0 +1,4 @@
+export declare const command = "markdown <file>";
+export declare const desc = "Generate markdown files";
+export { builder } from "./builder.js";
+export { handler } from "./handler.js";

package/lib/cli/commands/markdown/index.js

@@ -0,0 +1,4 @@
+export const command = "markdown <file>";
+export const desc = "Generate markdown files";
+export { builder } from "./builder.js";
+export { handler } from "./handler.js";

package/lib/methods/mulmo_beat.d.ts

@@ -1,6 +1,11 @@
 import { MulmoBeat } from "../types/index.js";
 export declare const MulmoBeatMethods: {
     getHtmlPrompt(beat: MulmoBeat): string | undefined;
-    getPlugin(beat: MulmoBeat): typeof import("../utils/image_plugins/text_slide.js") | typeof import("../utils/image_plugins/markdown.js") | typeof import("../utils/image_plugins/chart.js") | typeof import("../utils/image_plugins/mermaid.js") | typeof import("../utils/image_plugins/html_tailwind.js") | typeof import("../utils/image_plugins/image.js") | typeof import("../utils/image_plugins/movie.js") | typeof import("../utils/image_plugins/beat.js") | typeof import("../utils/image_plugins/voice_over.js") | typeof import("../utils/image_plugins/vision.js");
+    getPlugin(beat: MulmoBeat): {
+        imageType: string;
+        process: (params: import("../types/type.js").ImageProcessorParams) => Promise<string | undefined> | void;
+        path: (params: import("../types/type.js").ImageProcessorParams) => string | undefined;
+        markdown?: (params: import("../types/type.js").ImageProcessorParams) => string | undefined;
+    };
     getImageReferenceForImageGenerator(beat: MulmoBeat, imageRefs: Record<string, string>): string[];
 };

package/lib/types/schema.d.ts

@@ -5777,8 +5777,12 @@ export declare const mulmoStudioBeatSchema: z.ZodObject<{
     lipSyncFile: z.ZodOptional<z.ZodString>;
     captionFile: z.ZodOptional<z.ZodString>;
     htmlImageFile: z.ZodOptional<z.ZodString>;
+    markdown: z.ZodOptional<z.ZodString>;
+    html: z.ZodOptional<z.ZodString>;
 }, "strict", z.ZodTypeAny, {
     duration?: number | undefined;
+    markdown?: string | undefined;
+    html?: string | undefined;
     id?: string | undefined;
     startAt?: number | undefined;
     hash?: string | undefined;
@@ -5795,6 +5799,8 @@ export declare const mulmoStudioBeatSchema: z.ZodObject<{
     htmlImageFile?: string | undefined;
 }, {
     duration?: number | undefined;
+    markdown?: string | undefined;
+    html?: string | undefined;
     id?: string | undefined;
     startAt?: number | undefined;
     hash?: string | undefined;
@@ -6027,18 +6033,24 @@ export declare const mulmoSessionStateSchema: z.ZodObject<{
         multiLingual: z.ZodBoolean;
         caption: z.ZodBoolean;
         pdf: z.ZodBoolean;
+        markdown: z.ZodBoolean;
+        html: z.ZodBoolean;
     }, "strip", z.ZodTypeAny, {
         image: boolean;
         video: boolean;
         audio: boolean;
+        markdown: boolean;
         pdf: boolean;
+        html: boolean;
         multiLingual: boolean;
         caption: boolean;
     }, {
         image: boolean;
         video: boolean;
         audio: boolean;
+        markdown: boolean;
         pdf: boolean;
+        html: boolean;
         multiLingual: boolean;
         caption: boolean;
     }>;
@@ -6078,7 +6090,9 @@ export declare const mulmoSessionStateSchema: z.ZodObject<{
         image: boolean;
         video: boolean;
         audio: boolean;
+        markdown: boolean;
         pdf: boolean;
+        html: boolean;
         multiLingual: boolean;
         caption: boolean;
     };
@@ -6098,7 +6112,9 @@ export declare const mulmoSessionStateSchema: z.ZodObject<{
         image: boolean;
         video: boolean;
         audio: boolean;
+        markdown: boolean;
         pdf: boolean;
+        html: boolean;
         multiLingual: boolean;
         caption: boolean;
     };
@@ -8479,8 +8495,12 @@ export declare const mulmoStudioSchema: z.ZodObject<{
         lipSyncFile: z.ZodOptional<z.ZodString>;
         captionFile: z.ZodOptional<z.ZodString>;
         htmlImageFile: z.ZodOptional<z.ZodString>;
+        markdown: z.ZodOptional<z.ZodString>;
+        html: z.ZodOptional<z.ZodString>;
     }, "strict", z.ZodTypeAny, {
         duration?: number | undefined;
+        markdown?: string | undefined;
+        html?: string | undefined;
         id?: string | undefined;
         startAt?: number | undefined;
         hash?: string | undefined;
@@ -8497,6 +8517,8 @@ export declare const mulmoStudioSchema: z.ZodObject<{
         htmlImageFile?: string | undefined;
     }, {
         duration?: number | undefined;
+        markdown?: string | undefined;
+        html?: string | undefined;
         id?: string | undefined;
         startAt?: number | undefined;
         hash?: string | undefined;
@@ -8515,6 +8537,8 @@ export declare const mulmoStudioSchema: z.ZodObject<{
 }, "strict", z.ZodTypeAny, {
     beats: {
         duration?: number | undefined;
+        markdown?: string | undefined;
+        html?: string | undefined;
         id?: string | undefined;
         startAt?: number | undefined;
         hash?: string | undefined;
@@ -8862,6 +8886,8 @@ export declare const mulmoStudioSchema: z.ZodObject<{
 }, {
     beats: {
         duration?: number | undefined;
+        markdown?: string | undefined;
+        html?: string | undefined;
         id?: string | undefined;
         startAt?: number | undefined;
         hash?: string | undefined;

package/lib/types/schema.js

@@ -404,6 +404,8 @@ export const mulmoStudioBeatSchema = z
     lipSyncFile: z.string().optional(), // path to the lip sync file
     captionFile: z.string().optional(), // path to the caption image
     htmlImageFile: z.string().optional(), // path to the html image
+    markdown: z.string().optional(), // markdown string (alternative to image)
+    html: z.string().optional(), // html string (alternative to image)
 })
     .strict();
 export const mulmoStudioMultiLingualDataSchema = z.object({
@@ -424,6 +426,8 @@ export const mulmoSessionStateSchema = z.object({
         multiLingual: z.boolean(),
         caption: z.boolean(),
         pdf: z.boolean(),
+        markdown: z.boolean(),
+        html: z.boolean(),
     }),
     inBeatSession: z.object({
         audio: z.record(z.string(), z.boolean()),

package/lib/types/type.d.ts

@@ -85,7 +85,7 @@ export type Text2HtmlAgentInfo = {
 };
 export type BeatMediaType = "movie" | "image";
 export type StoryToScriptGenerateMode = (typeof storyToScriptGenerateMode)[keyof typeof storyToScriptGenerateMode];
-export type SessionType = "audio" | "image" | "video" | "multiLingual" | "caption" | "pdf";
+export type SessionType = "audio" | "image" | "video" | "multiLingual" | "caption" | "pdf" | "markdown" | "html";
 export type BeatSessionType = "audio" | "image" | "multiLingual" | "caption" | "movie" | "html" | "imageReference" | "soundEffect" | "lipSync";
 export type SessionProgressEvent = {
     kind: "session";

package/lib/utils/context.d.ts

@@ -2,6 +2,8 @@ import type { MulmoStudioBeat, MulmoScript, MulmoPresentationStyle, MulmoStudioM
 export declare const createStudioData: (_mulmoScript: MulmoScript, fileName: string, videoCaptionLang?: string, presentationStyle?: MulmoPresentationStyle | null) => {
     beats: {
         duration?: number | undefined;
+        markdown?: string | undefined;
+        html?: string | undefined;
         id?: string | undefined;
         startAt?: number | undefined;
         hash?: string | undefined;
@@ -354,6 +356,8 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
     studio: {
         beats: {
             duration?: number | undefined;
+            markdown?: string | undefined;
+            html?: string | undefined;
             id?: string | undefined;
             startAt?: number | undefined;
             hash?: string | undefined;
@@ -831,6 +835,8 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
             multiLingual: boolean;
             caption: boolean;
             pdf: boolean;
+            markdown: boolean;
+            html: boolean;
         };
         inBeatSession: {
             audio: {};

package/lib/utils/context.js

@@ -34,6 +34,8 @@ const initSessionState = () => {
             multiLingual: false,
             caption: false,
             pdf: false,
+            markdown: false,
+            html: false,
         },
         inBeatSession: {
             audio: {},

package/lib/utils/image_plugins/index.d.ts

@@ -1,11 +1,7 @@
-import * as pluginTextSlide from "./text_slide.js";
-import * as pluginMarkdown from "./markdown.js";
-import * as pluginChart from "./chart.js";
-import * as pluginMermaid from "./mermaid.js";
-import * as pluginHtmlTailwind from "./html_tailwind.js";
-import * as pluginImage from "./image.js";
-import * as pluginMovie from "./movie.js";
-import * as pluginBeat from "./beat.js";
-import * as pluginVoiceOver from "./voice_over.js";
-import * as pluginVision from "./vision.js";
-export declare const findImagePlugin: (imageType?: string) => typeof pluginTextSlide | typeof pluginMarkdown | typeof pluginChart | typeof pluginMermaid | typeof pluginHtmlTailwind | typeof pluginImage | typeof pluginMovie | typeof pluginBeat | typeof pluginVoiceOver | typeof pluginVision | undefined;
+import { ImageProcessorParams } from "../../types/index.js";
+export declare const findImagePlugin: (imageType?: string) => {
+    imageType: string;
+    process: (params: ImageProcessorParams) => Promise<string | undefined> | void;
+    path: (params: ImageProcessorParams) => string | undefined;
+    markdown?: (params: ImageProcessorParams) => string | undefined;
+} | undefined;

package/lib/utils/image_plugins/index.js

@@ -8,18 +8,7 @@ import * as pluginMovie from "./movie.js";
 import * as pluginBeat from "./beat.js";
 import * as pluginVoiceOver from "./voice_over.js";
 import * as pluginVision from "./vision.js";
-const imagePlugins = [
-    pluginTextSlide,
-    pluginMarkdown,
-    pluginImage,
-    pluginChart,
-    pluginMermaid,
-    pluginMovie,
-    pluginHtmlTailwind,
-    pluginBeat,
-    pluginVoiceOver,
-    pluginVision,
-];
+const imagePlugins = [pluginTextSlide, pluginMarkdown, pluginImage, pluginChart, pluginMermaid, pluginMovie, pluginHtmlTailwind, pluginBeat, pluginVoiceOver, pluginVision];
 export const findImagePlugin = (imageType) => {
     return imagePlugins.find((plugin) => plugin.imageType === imageType);
 };

package/lib/utils/image_plugins/markdown.d.ts

@@ -2,3 +2,4 @@ import { ImageProcessorParams } from "../../types/index.js";
 export declare const imageType = "markdown";
 export declare const process: (params: ImageProcessorParams) => Promise<string | undefined>;
 export declare const path: (params: ImageProcessorParams) => string;
+export declare const markdown: (params: ImageProcessorParams) => string | undefined;

package/lib/utils/image_plugins/markdown.js

@@ -5,9 +5,16 @@ const processMarkdown = async (params) => {
     const { beat, imagePath, textSlideStyle, canvasSize } = params;
     if (!beat.image || beat.image.type !== imageType)
         return;
-    const markdown = Array.isArray(beat.image.markdown) ? beat.image.markdown.join("\n") : beat.image.markdown;
+    const markdown = dumpMarkdown(params) ?? "";
     await renderMarkdownToImage(markdown, textSlideStyle, imagePath, canvasSize.width, canvasSize.height);
     return imagePath;
 };
+const dumpMarkdown = (params) => {
+    const { beat } = params;
+    if (!beat.image || beat.image.type !== imageType)
+        return;
+    return Array.isArray(beat.image.markdown) ? beat.image.markdown.join("\n") : beat.image.markdown;
+};
 export const process = processMarkdown;
 export const path = parrotingImagePath;
+export const markdown = dumpMarkdown;

package/lib/utils/image_plugins/mermaid.d.ts

@@ -2,3 +2,4 @@ import { ImageProcessorParams } from "../../types/index.js";
 export declare const imageType = "mermaid";
 export declare const process: (params: ImageProcessorParams) => Promise<string | undefined>;
 export declare const path: (params: ImageProcessorParams) => string;
+export declare const markdown: (params: ImageProcessorParams) => string | undefined;

package/lib/utils/image_plugins/mermaid.js

@@ -19,5 +19,14 @@ const processMermaid = async (params) => {
     }
     return imagePath;
 };
+const dumpMarkdown = (params) => {
+    const { beat } = params;
+    if (!beat.image || beat.image.type !== imageType)
+        return;
+    if (beat.image.code.kind !== "text")
+        return; // support only text for now
+    return `\`\`\`mermaid\n${beat.image.code.text}\n\`\`\``;
+};
 export const process = processMermaid;
 export const path = parrotingImagePath;
+export const markdown = dumpMarkdown;

package/lib/utils/image_plugins/text_slide.d.ts

@@ -2,3 +2,4 @@ import { ImageProcessorParams } from "../../types/index.js";
 export declare const imageType = "textSlide";
 export declare const process: (params: ImageProcessorParams) => Promise<string | undefined>;
 export declare const path: (params: ImageProcessorParams) => string;
+export declare const markdown: (params: ImageProcessorParams) => string | undefined;

package/lib/utils/image_plugins/text_slide.js

@@ -6,7 +6,7 @@ const processTextSlide = async (params) => {
     if (!beat.image || beat.image.type !== imageType)
         return;
     const slide = beat.image.slide;
-    const markdown = `# ${slide.title}\n` + (slide.subtitle ? `## ${slide.subtitle}\n` : "") + (slide.bullets ?? []).map((text) => `- ${text}`).join("\n");
+    const markdown = dumpMarkdown(params) ?? "";
     const topMargin = (() => {
         if (slide.bullets?.length && slide.bullets.length > 0) {
             return "";
@@ -17,5 +17,16 @@ const processTextSlide = async (params) => {
     await renderMarkdownToImage(markdown, textSlideStyle + topMargin, imagePath, canvasSize.width, canvasSize.height);
     return imagePath;
 };
+const dumpMarkdown = (params) => {
+    const { beat } = params;
+    if (!beat.image || beat.image.type !== imageType)
+        return;
+    const slide = beat.image.slide;
+    const titleString = slide.title ? `# ${slide.title}\n` : "";
+    const subtitleString = slide.subtitle ? `## ${slide.subtitle}\n` : "";
+    const bulletsString = (slide.bullets ?? []).map((text) => `- ${text}`).join("\n");
+    return `${titleString}${subtitleString}${bulletsString}`;
+};
 export const process = processTextSlide;
 export const path = parrotingImagePath;
+export const markdown = dumpMarkdown;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mulmocast",
-  "version": "1.2.28",
+  "version": "1.2.30",
   "description": "",
   "type": "module",
   "main": "lib/index.node.js",
@@ -49,6 +49,8 @@
     "scripting": "npx tsx ./src/cli/bin.ts tool scripting",
     "prompt": "npx tsx ./src/cli/bin.ts tool prompt",
     "schema": "npx tsx ./src/cli/bin.ts tool schema",
+    "markdown": "npx tsx ./src/cli/bin.ts markdown",
+    "html": "npx tsx ./src/cli/bin.ts html",
     "story_to_script": "npx tsx ./src/cli/bin.ts tool story_to_script",
     "whisper": "npx tsx ./src/cli/bin.ts tool whisper",
     "latest": "yarn upgrade-interactive  --latest",
@@ -67,7 +69,7 @@
   "homepage": "https://github.com/receptron/mulmocast-cli#readme",
   "dependencies": {
     "@google-cloud/text-to-speech": "^6.3.0",
-    "@google/genai": "^1.17.0",
+    "@google/genai": "^1.19.0",
     "@graphai/anthropic_agent": "^2.0.11",
     "@graphai/browserless_agent": "^2.0.1",
     "@graphai/gemini_agent": "^2.0.1",
@@ -77,20 +79,20 @@
     "@graphai/stream_agent_filter": "^2.0.2",
     "@graphai/vanilla": "^2.0.12",
     "@graphai/vanilla_node_agents": "^2.0.4",
-    "@inquirer/input": "^4.2.2",
-    "@inquirer/select": "^4.3.2",
-    "@modelcontextprotocol/sdk": "^1.17.5",
+    "@inquirer/input": "^4.2.4",
+    "@inquirer/select": "^4.3.4",
+    "@modelcontextprotocol/sdk": "^1.18.0",
     "@mozilla/readability": "^0.6.0",
     "@tavily/core": "^0.5.11",
     "clipboardy": "^4.0.0",
     "dotenv": "^17.2.2",
     "fluent-ffmpeg": "^2.1.3",
     "graphai": "^2.0.14",
-    "jsdom": "^26.1.0",
-    "marked": "^16.2.1",
-    "mulmocast-vision": "^0.1.1",
+    "jsdom": "^27.0.0",
+    "marked": "^16.3.0",
+    "mulmocast-vision": "^1.0.2",
     "ora": "^8.2.0",
-    "puppeteer": "^24.19.0",
+    "puppeteer": "^24.20.0",
     "replicate": "^1.1.0",
     "yaml": "^2.8.1",
     "yargs": "^18.0.0",

package/scripts/test/test_image_refs.json

@@ -55,6 +55,16 @@
         "provider": "google",
         "model": "gemini-2.5-flash-image-preview"
       }
+    },
+    {
+      "id": "seedream-4",
+      "text": "Hello World with a witch and a broom with Seadream-4",
+      "imagePrompt": "Saying hello to the world",
+      "imageNames": ["witch", "broom"],
+      "imageParams": {
+        "provider": "replicate",
+        "model": "bytedance/seedream-4"
+      }
     }
   ]
 }

package/scripts/test/test_markdown.json

@@ -0,0 +1,60 @@
+{
+  "$mulmocast": {
+    "version": "1.1"
+  },
+  "lang": "en",
+  "title": "Test Markdown",
+  "description": "This is a test markdown file.",
+  "beats": [
+    {
+      "text": "Hello World",
+      "image": {
+        "type": "markdown",
+        "markdown": []
+      }
+    },
+    {
+      "text": "Hello World",
+      "image": {
+        "type": "markdown",
+        "markdown": ["## Chapter 1"]
+      }
+    },
+    {
+      "image": {
+        "type": "markdown",
+        "markdown": ["## Chapter 2", "- Hello", "- World"]
+      }
+    },
+    {
+      "image": {
+        "type": "textSlide",
+        "slide": {
+          "title": "Chapter 3",
+          "subtitle": "Subtitle",
+          "bullets": ["Hello", "World"]
+        }
+      }
+    },
+    {
+      "image": {
+        "type": "textSlide",
+        "slide": {
+          "title": "",
+          "subtitle": "Chapter 4",
+          "bullets": ["Hello", "World"]
+        }
+      }
+    },
+    {
+      "image": {
+        "type": "mermaid",
+        "title": "Business Process Flow",
+        "code": {
+          "kind": "text",
+          "text": "graph LR\n    A[Market Research] --> B[Product Planning]\n    B --> C[Development]\n    C --> D[Testing]\n    D --> E[Manufacturing]\n    E --> F[Marketing]\n    F --> G[Sales]\n    G --> H[Customer Support]\n    H --> A"
+        }
+      }
+    }
+  ]
+}