mulmocast 0.0.22 → 0.0.23

package/lib/types/schema.js

@@ -84,6 +84,18 @@ export const mulmoTextSlideMediaSchema = z
     }),
 })
     .strict();
+export const mulmoCaptionParamsSchema = z
+    .object({
+    lang: langSchema.optional(),
+    textColor: z.string().default("#FFFFFF"),
+    backgroundColor: z.string().default("#000000"),
+    fontSize: z.number().default(16),
+    fontFamily: z.string().default("Arial"),
+    fontWeight: z.string().default("normal"),
+    textAlign: z.string().default("left"),
+    textDecoration: z.string().default("none"),
+})
+    .strict();
 export const mulmoChartMediaSchema = z
     .object({
     type: z.literal("chart"),
@@ -141,6 +153,11 @@ const mulmoMidiMediaSchema = z
 export const mulmoAudioAssetSchema = z.union([mulmoAudioMediaSchema, mulmoMidiMediaSchema]);
 const imageIdSchema = z.string();
 export const mulmoImageParamsImagesSchema = z.record(imageIdSchema, mulmoImageMediaSchema);
+export const mulmoFillOptionSchema = z
+    .object({
+    style: z.enum(["aspectFit", "aspectFill"]).default("aspectFit"),
+})
+    .describe("How to handle aspect ratio differences between image and canvas");
 export const mulmoImageParamsSchema = z
     .object({
     model: z.string().optional(), // default: provider specific
@@ -159,6 +176,11 @@ export const beatAudioParamsSchema = z
     padding: z.number().optional().describe("Padding between beats"), // seconds
 })
     .strict();
+export const mulmoHtmlImageParamsSchema = z
+    .object({
+    model: z.string().optional(), // default: provider specific
+})
+    .strict();
 // Note: we can't extend beatAudioParamsSchema because it has padding as optional
 export const audioParamsSchema = z
     .object({
@@ -190,8 +212,15 @@ export const mulmoBeatSchema = z
     duration: z.number().optional().describe("Duration of the beat. Used only when the text is empty"),
     imageParams: mulmoImageParamsSchema.optional(), // beat specific parameters
     audioParams: beatAudioParamsSchema.optional(), // beat specific parameters
+    movieParams: z
+        .object({
+        fillOption: mulmoFillOptionSchema,
+    })
+        .optional(),
+    htmlImageParams: mulmoHtmlImageParamsSchema.optional(),
     speechOptions: speechOptionsSchema.optional(),
     textSlideParams: textSlideParamsSchema.optional(),
+    captionParams: mulmoCaptionParamsSchema.optional(),
     imageNames: z.array(imageIdSchema).optional(), // list of image names to use for image generation. The default is all images in the imageParams.images.
     imagePrompt: z.string().optional(),
     moviePrompt: z.string().optional(),
@@ -218,7 +247,8 @@ export const mulmoSpeechParamsSchema = z
 })
     .strict();
 export const text2ImageProviderSchema = z.union([z.literal("openai"), z.literal("google")]).default("openai");
-export const text2MovieProviderSchema = z.union([z.literal("openai"), z.literal("google")]).default("google");
+export const text2HtmlImageProviderSchema = z.union([z.literal("openai"), z.literal("anthropic")]).default("openai");
+export const text2MovieProviderSchema = z.union([z.literal("openai"), z.literal("google"), z.literal("replicate")]).default("google");
 export const mulmoTransitionSchema = z.object({
     type: z.enum(["fade", "slideout_left"]),
     duration: z.number().min(0).max(2).default(0.3), // transition duration in seconds
@@ -228,6 +258,7 @@ export const mulmoMovieParamsSchema = z
     provider: text2MovieProviderSchema.optional(),
     model: z.string().optional(), // default: provider specific
     transition: mulmoTransitionSchema.optional(),
+    fillOption: mulmoFillOptionSchema.optional(),
 })
     .strict();
 export const mulmoPresentationStyleSchema = z.object({
@@ -249,8 +280,14 @@ export const mulmoPresentationStyleSchema = z.object({
     })
         .optional(),
     movieParams: mulmoMovieParamsSchema.optional(),
+    htmlImageParams: mulmoHtmlImageParamsSchema
+        .extend({
+        provider: text2HtmlImageProviderSchema,
+    })
+        .optional(),
     // for textSlides
     textSlideParams: textSlideParamsSchema.optional(),
+    captionParams: mulmoCaptionParamsSchema.optional(),
     audioParams: audioParamsSchema.default({
         introPadding: 1.0,
         padding: 0.3,

package/lib/types/type.d.ts

@@ -1,4 +1,4 @@
-import { langSchema, localizedTextSchema, mulmoBeatSchema, mulmoScriptSchema, mulmoStudioSchema, mulmoStudioBeatSchema, mulmoStoryboardSchema, mulmoStoryboardSceneSchema, mulmoStudioMultiLingualSchema, mulmoStudioMultiLingualDataSchema, speakerDictionarySchema, mulmoImageParamsSchema, mulmoImageParamsImagesSchema, mulmoMovieParamsSchema, mulmoSpeechParamsSchema, textSlideParamsSchema, speechOptionsSchema, speakerDataSchema, mulmoCanvasDimensionSchema, mulmoScriptTemplateSchema, mulmoScriptTemplateFileSchema, text2ImageProviderSchema, text2MovieProviderSchema, text2SpeechProviderSchema, mulmoPresentationStyleSchema, multiLingualTextsSchema, mulmoMermaidMediaSchema, mulmoTextSlideMediaSchema, mulmoMarkdownMediaSchema, mulmoImageMediaSchema, mulmoChartMediaSchema, mediaSourceSchema, mulmoSessionStateSchema } from "./schema.js";
+import { langSchema, localizedTextSchema, mulmoBeatSchema, mulmoScriptSchema, mulmoStudioSchema, mulmoStudioBeatSchema, mulmoStoryboardSchema, mulmoStoryboardSceneSchema, mulmoStudioMultiLingualSchema, mulmoStudioMultiLingualDataSchema, speakerDictionarySchema, mulmoImageParamsSchema, mulmoImageParamsImagesSchema, mulmoFillOptionSchema, mulmoMovieParamsSchema, mulmoSpeechParamsSchema, textSlideParamsSchema, speechOptionsSchema, speakerDataSchema, mulmoCanvasDimensionSchema, mulmoScriptTemplateSchema, mulmoScriptTemplateFileSchema, text2ImageProviderSchema, text2HtmlImageProviderSchema, text2MovieProviderSchema, text2SpeechProviderSchema, mulmoPresentationStyleSchema, multiLingualTextsSchema, mulmoMermaidMediaSchema, mulmoTextSlideMediaSchema, mulmoMarkdownMediaSchema, mulmoImageMediaSchema, mulmoChartMediaSchema, mediaSourceSchema, mulmoSessionStateSchema } from "./schema.js";
 import { pdf_modes, pdf_sizes, storyToScriptGenerateMode } from "../utils/const.js";
 import { LLM } from "../utils/utils.js";
 import { z } from "zod";
@@ -10,8 +10,10 @@ export type SpeechOptions = z.infer<typeof speechOptionsSchema>;
 export type SpeakerData = z.infer<typeof speakerDataSchema>;
 export type MulmoImageParams = z.infer<typeof mulmoImageParamsSchema>;
 export type MulmoImageParamsImages = z.infer<typeof mulmoImageParamsImagesSchema>;
+export type MulmoFillOption = z.infer<typeof mulmoFillOptionSchema>;
 export type TextSlideParams = z.infer<typeof textSlideParamsSchema>;
 export type Text2ImageProvider = z.infer<typeof text2ImageProviderSchema>;
+export type Text2HtmlImageProvider = z.infer<typeof text2HtmlImageProviderSchema>;
 export type Text2MovieProvider = z.infer<typeof text2MovieProviderSchema>;
 export type Text2SpeechProvider = z.infer<typeof text2SpeechProviderSchema>;
 export type LocalizedText = z.infer<typeof localizedTextSchema>;
@@ -49,7 +51,6 @@ export type MulmoStudioContext = {
     lang?: string;
     dryRun?: boolean;
     force: boolean;
-    caption?: string;
     sessionState: MulmoSessionState;
     presentationStyle: MulmoPresentationStyle;
     multiLingual: MulmoStudioMultiLingual;
@@ -62,6 +63,7 @@ export type ScriptingParams = {
     filename: string;
     llm_model?: string;
     llm?: LLM;
+    verbose?: boolean;
 };
 export type ImageProcessorParams = {
     beat: MulmoBeat;
@@ -77,6 +79,11 @@ export type Text2ImageAgentInfo = {
     agent: string;
     imageParams: MulmoImageParams;
 };
+export type Text2HtmlAgentInfo = {
+    provider: Text2HtmlImageProvider;
+    agent: string;
+    model: string;
+};
 export type BeatMediaType = "movie" | "image";
 export type StoryToScriptGenerateMode = (typeof storyToScriptGenerateMode)[keyof typeof storyToScriptGenerateMode];
 export type SessionType = "audio" | "image" | "video" | "multiLingual" | "caption" | "pdf";

package/lib/utils/ffmpeg_utils.d.ts

@@ -7,7 +7,7 @@ export type FfmpegContext = {
 export declare const FfmpegContextInit: () => FfmpegContext;
 export declare const FfmpegContextAddInput: (context: FfmpegContext, input: string, inputOptions?: string[]) => number;
 export declare const FfmpegContextPushFormattedAudio: (context: FfmpegContext, sourceId: string, outputId: string, duration?: number | undefined) => void;
-export declare const FfmpegContextInputFormattedAudio: (context: FfmpegContext, input: string, duration?: number | undefined) => string;
+export declare const FfmpegContextInputFormattedAudio: (context: FfmpegContext, input: string, duration?: number | undefined, inputOptions?: string[]) => string;
 export declare const FfmpegContextGenerateOutput: (context: FfmpegContext, output: string, options?: string[]) => Promise<number>;
 export declare const ffmpegGetMediaDuration: (filePath: string) => Promise<number>;
 export declare const extractImageFromMovie: (movieFile: string, imagePath: string) => Promise<void>;

package/lib/utils/ffmpeg_utils.js

@@ -25,8 +25,8 @@ export const FfmpegContextPushFormattedAudio = (context, sourceId, outputId, dur
         context.filterComplex.push(`${sourceId}aformat=sample_fmts=fltp:sample_rates=44100:channel_layouts=stereo${outputId}`);
     }
 };
-export const FfmpegContextInputFormattedAudio = (context, input, duration = undefined) => {
-    const index = FfmpegContextAddInput(context, input);
+export const FfmpegContextInputFormattedAudio = (context, input, duration = undefined, inputOptions) => {
+    const index = FfmpegContextAddInput(context, input, inputOptions);
     const audioId = `[a${index}]`;
     FfmpegContextPushFormattedAudio(context, `[${index}:a]`, audioId, duration);
     return audioId;

package/lib/utils/preprocess.d.ts

@@ -182,8 +182,8 @@ export declare const createOrUpdateStudioData: (_mulmoScript: MulmoScript, curre
             } | undefined;
             description?: string | undefined;
             imageParams?: {
-                model?: string | undefined;
                 style?: string | undefined;
+                model?: string | undefined;
                 moderation?: string | undefined;
                 images?: Record<string, {
                     type: "image";
@@ -205,9 +205,27 @@ export declare const createOrUpdateStudioData: (_mulmoScript: MulmoScript, curre
             audioParams?: {
                 padding?: number | undefined;
             } | undefined;
+            movieParams?: {
+                fillOption: {
+                    style: "aspectFit" | "aspectFill";
+                };
+            } | undefined;
+            htmlImageParams?: {
+                model?: string | undefined;
+            } | undefined;
             textSlideParams?: {
                 cssStyles: string | string[];
             } | undefined;
+            captionParams?: {
+                textColor: string;
+                backgroundColor: string;
+                fontSize: number;
+                fontFamily: string;
+                fontWeight: string;
+                textAlign: string;
+                textDecoration: string;
+                lang?: string | undefined;
+            } | undefined;
             imageNames?: string[] | undefined;
             imagePrompt?: string | undefined;
             moviePrompt?: string | undefined;
@@ -223,8 +241,8 @@ export declare const createOrUpdateStudioData: (_mulmoScript: MulmoScript, curre
         description?: string | undefined;
         imageParams?: {
             provider: "openai" | "google";
-            model?: string | undefined;
             style?: string | undefined;
+            model?: string | undefined;
             moderation?: string | undefined;
             images?: Record<string, {
                 type: "image";
@@ -243,17 +261,34 @@ export declare const createOrUpdateStudioData: (_mulmoScript: MulmoScript, curre
                 };
             }> | undefined;
         } | undefined;
-        textSlideParams?: {
-            cssStyles: string | string[];
-        } | undefined;
         movieParams?: {
-            provider?: "openai" | "google" | undefined;
+            provider?: "openai" | "google" | "replicate" | undefined;
             model?: string | undefined;
+            fillOption?: {
+                style: "aspectFit" | "aspectFill";
+            } | undefined;
             transition?: {
                 type: "fade" | "slideout_left";
                 duration: number;
             } | undefined;
         } | undefined;
+        htmlImageParams?: {
+            provider: "openai" | "anthropic";
+            model?: string | undefined;
+        } | undefined;
+        textSlideParams?: {
+            cssStyles: string | string[];
+        } | undefined;
+        captionParams?: {
+            textColor: string;
+            backgroundColor: string;
+            fontSize: number;
+            fontFamily: string;
+            fontWeight: string;
+            textAlign: string;
+            textDecoration: string;
+            lang?: string | undefined;
+        } | undefined;
         references?: {
             type: "image" | "audio" | "article" | "paper" | "video";
             url: string;

package/lib/utils/utils.d.ts

@@ -16,3 +16,4 @@ export declare const isHttp: (fileOrUrl: string) => boolean;
 export declare const text2hash: (input: string) => string;
 export declare const localizedText: (beat: MulmoBeat, multiLingualData?: MulmoStudioMultiLingualData, lang?: string) => string;
 export declare const sleep: (milliseconds: number) => Promise<unknown>;
+export declare function userAssert(condition: boolean, message: string): asserts condition;

package/lib/utils/utils.js

@@ -55,3 +55,8 @@ export const localizedText = (beat, multiLingualData, lang) => {
 export const sleep = async (milliseconds) => {
     return await new Promise((resolve) => setTimeout(resolve, milliseconds));
 };
+export function userAssert(condition, message) {
+    if (!condition) {
+        throw new Error(message);
+    }
+}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mulmocast",
-  "version": "0.0.22",
+  "version": "0.0.23",
   "description": "",
   "type": "module",
   "main": "lib/index.js",
@@ -74,11 +74,12 @@
     "dotenv": "^16.4.7",
     "fluent-ffmpeg": "^2.1.3",
     "google-auth-library": "^9.15.1",
-    "graphai": "^2.0.8",
+    "graphai": "^2.0.9",
     "inquirer": "^12.6.1",
     "marked": "^15.0.12",
     "ora": "^8.2.0",
     "puppeteer": "^24.10.2",
+    "replicate": "^1.0.1",
     "yaml": "^2.8.0",
     "yargs": "^17.7.2",
     "zod": "^3.25.67",

package/scripts/templates/presentation.json

@@ -0,0 +1,123 @@
+{
+  "$mulmocast": {
+    "version": "1.0",
+    "credit": "closing"
+  },
+  "htmlImageParams": {
+    "provider": "anthropic",
+    "model": "claude-3-7-sonnet-20250219"
+  },
+  "title": "Sample Title",
+  "references": [
+    {
+      "url": "https://www.somegreatwebsite.com/article/123",
+      "title": "Title of the article we are referencing",
+      "type": "article"
+    }
+  ],
+  "lang": "en",
+  "beats": [
+    {
+      "speaker": "Presenter",
+      "text": "Book rooms with locals, not hotels.",
+      "htmlPrompt": {
+        "prompt": "Create a clean and bold title slide featuring the original name 'AirBed & Breakfast' with the tagline: 'Book rooms with locals, not hotels'. Include the old logo and a travel-themed background."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Hotels are expensive and detached from local culture.",
+      "htmlPrompt": {
+        "prompt": "Design a slide that lists the main problems with hotels: high prices, lack of personality, and cultural disconnection. Use simple icons for cost, generic room, and traveler."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Airbnb lets people rent out their spaces, offering affordable, authentic experiences.",
+      "htmlPrompt": {
+        "prompt": "Illustrate Airbnb's solution: locals listing their spare rooms or homes, travelers booking directly, and both benefiting. Use a flow diagram to represent this."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Craigslist and Couchsurfing prove the demand for alternative lodging.",
+      "htmlPrompt": {
+        "prompt": "Show a bar graph comparing Craigslist listings, Couchsurfing users, and early Airbnb listings to demonstrate existing demand and opportunity."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "A massive market: over 630 million travel bookings annually.",
+      "htmlPrompt": {
+        "prompt": "Visualize the global travel booking market. Use a world map and large numbers to emphasize the size and growth potential."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Our platform enables hosts to list once and earn; guests search, book, and review.",
+      "htmlPrompt": {
+        "prompt": "Create a product overview slide showing how the platform works from host listing to guest review. Highlight simplicity and trust."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "We charge a 10% commission per booking.",
+      "htmlPrompt": {
+        "prompt": "Display a slide with the monetization strategy—Airbnb earns through a 10% commission. Use a simple pie chart or booking flow with fee annotation."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Competitors include hotels, Craigslist, Couchsurfing — we offer a better, scalable solution.",
+      "htmlPrompt": {
+        "prompt": "Design a comparison table: price, uniqueness, trust, scalability. Airbnb should clearly stand out on all dimensions."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Key advantages: lower price, wide selection, ease of use, host incentives, trusted system.",
+      "htmlPrompt": {
+        "prompt": "List Airbnb’s competitive advantages using 6 icons or badges, one for each feature."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Founders: Brian Chesky (Design), Joe Gebbia (Marketing), Nathan Blecharczyk (Engineering).",
+      "htmlPrompt": {
+        "prompt": "Introduce the founding team with photos and brief bios. Highlight their roles and strengths in product, marketing, and tech."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Media buzz: featured on TechCrunch and buzz from SXSW.",
+      "htmlPrompt": {
+        "prompt": "Showcase media recognition. Include TechCrunch logo and SXSW quotes or metrics. This adds credibility and traction."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Early users love the authentic and affordable experiences.",
+      "htmlPrompt": {
+        "prompt": "Add 2–3 short testimonials from early users (guests and hosts), displayed in speech bubbles with faces or usernames."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "We’re raising $600K to support 12 months of growth and reach 80K bookings with $2M in revenue.",
+      "htmlPrompt": {
+        "prompt": "Closing investment slide. Include fund usage (pie chart: 40% marketing, 30% product, 30% ops), and KPIs: 80K bookings, $2M revenue.",
+        "data": {
+          "ask": 600000,
+          "duration_months": 12,
+          "target_bookings": 80000,
+          "target_revenue": 2000000,
+          "use_of_funds": {
+            "marketing": 40,
+            "product": 30,
+            "operations": 30
+          }
+        }
+      }
+    }
+  ]
+}

package/scripts/templates/presentation.json~

@@ -0,0 +1,119 @@
+{
+  "$mulmocast": {
+    "version": "1.0",
+    "credit": "closing"
+  },
+  "title": "Sample Title",
+  "references": [
+    {
+      "url": "https://www.somegreatwebsite.com/article/123",
+      "title": "Title of the article we are referencing",
+      "type": "article"
+    }
+  ],
+  "lang": "en",
+  "beats": [
+    {
+      "speaker": "Presenter",
+      "text": "Book rooms with locals, not hotels.",
+      "htmlPrompt": {
+        "prompt": "Create a clean and bold title slide featuring the original name 'AirBed & Breakfast' with the tagline: 'Book rooms with locals, not hotels'. Include the old logo and a travel-themed background."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Hotels are expensive and detached from local culture.",
+      "htmlPrompt": {
+        "prompt": "Design a slide that lists the main problems with hotels: high prices, lack of personality, and cultural disconnection. Use simple icons for cost, generic room, and traveler."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Airbnb lets people rent out their spaces, offering affordable, authentic experiences.",
+      "htmlPrompt": {
+        "prompt": "Illustrate Airbnb's solution: locals listing their spare rooms or homes, travelers booking directly, and both benefiting. Use a flow diagram to represent this."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Craigslist and Couchsurfing prove the demand for alternative lodging.",
+      "htmlPrompt": {
+        "prompt": "Show a bar graph comparing Craigslist listings, Couchsurfing users, and early Airbnb listings to demonstrate existing demand and opportunity."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "A massive market: over 630 million travel bookings annually.",
+      "htmlPrompt": {
+        "prompt": "Visualize the global travel booking market. Use a world map and large numbers to emphasize the size and growth potential."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Our platform enables hosts to list once and earn; guests search, book, and review.",
+      "htmlPrompt": {
+        "prompt": "Create a product overview slide showing how the platform works from host listing to guest review. Highlight simplicity and trust."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "We charge a 10% commission per booking.",
+      "htmlPrompt": {
+        "prompt": "Display a slide with the monetization strategy—Airbnb earns through a 10% commission. Use a simple pie chart or booking flow with fee annotation."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Competitors include hotels, Craigslist, Couchsurfing — we offer a better, scalable solution.",
+      "htmlPrompt": {
+        "prompt": "Design a comparison table: price, uniqueness, trust, scalability. Airbnb should clearly stand out on all dimensions."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Key advantages: lower price, wide selection, ease of use, host incentives, trusted system.",
+      "htmlPrompt": {
+        "prompt": "List Airbnb’s competitive advantages using 6 icons or badges, one for each feature."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Founders: Brian Chesky (Design), Joe Gebbia (Marketing), Nathan Blecharczyk (Engineering).",
+      "htmlPrompt": {
+        "prompt": "Introduce the founding team with photos and brief bios. Highlight their roles and strengths in product, marketing, and tech."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Media buzz: featured on TechCrunch and buzz from SXSW.",
+      "htmlPrompt": {
+        "prompt": "Showcase media recognition. Include TechCrunch logo and SXSW quotes or metrics. This adds credibility and traction."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "Early users love the authentic and affordable experiences.",
+      "htmlPrompt": {
+        "prompt": "Add 2–3 short testimonials from early users (guests and hosts), displayed in speech bubbles with faces or usernames."
+      }
+    },
+    {
+      "speaker": "Presenter",
+      "text": "We’re raising $600K to support 12 months of growth and reach 80K bookings with $2M in revenue.",
+      "htmlPrompt": {
+        "prompt": "Closing investment slide. Include fund usage (pie chart: 40% marketing, 30% product, 30% ops), and KPIs: 80K bookings, $2M revenue.",
+        "data": {
+          "ask": 600000,
+          "duration_months": 12,
+          "target_bookings": 80000,
+          "target_revenue": 2000000,
+          "use_of_funds": {
+            "marketing": 40,
+            "product": 30,
+            "operations": 30
+          }
+        }
+      }
+    }
+  ]
+}