@mindstudio-ai/remy 0.1.42 → 0.1.44

package/dist/headless.js

@@ -2225,7 +2225,7 @@ function startStatusWatcher(config) {
     inflight = true;
     try {
       const ctx = getContext();
-      if (!ctx.assistantText && !ctx.lastToolName) {
+      if (!ctx.assistantText && !ctx.lastToolName && !ctx.userMessage) {
         return;
       }
       const res = await fetch(url, {
@@ -2368,7 +2368,8 @@ ${partial}` : "[INTERRUPTED] Agent was interrupted before producing output.",
         getContext: () => ({
           assistantText: getPartialText(contentBlocks),
           lastToolName: currentToolNames || void 0,
-          lastToolResult: lastToolResult || void 0
+          lastToolResult: lastToolResult || void 0,
+          userMessage: task
         }),
         onStatus: (label) => emit2({ type: "status", message: label }),
         signal
@@ -2597,14 +2598,30 @@ Current date/time: ${(/* @__PURE__ */ new Date()).toISOString().replace("T", " "
     return wrapRun();
   }
   log4.info("Sub-agent backgrounded", { requestId, parentToolId, agentName });
+  toolRegistry?.register({
+    id: parentToolId,
+    name: agentName,
+    input: { task },
+    abortController: bgAbort,
+    startedAt: Date.now(),
+    settle: () => {
+    },
+    rerun: () => {
+    },
+    getPartialResult: () => ""
+  });
   const ack = await generateBackgroundAck({
     apiConfig,
     agentName: subAgentId || "agent",
     task
   });
-  wrapRun().then((finalResult) => onBackgroundComplete?.(finalResult)).catch(
-    (err) => onBackgroundComplete?.({ text: `Error: ${err.message}`, messages: [] })
-  );
+  wrapRun().then((finalResult) => {
+    toolRegistry?.unregister(parentToolId);
+    onBackgroundComplete?.(finalResult);
+  }).catch((err) => {
+    toolRegistry?.unregister(parentToolId);
+    onBackgroundComplete?.({ text: `Error: ${err.message}`, messages: [] });
+  });
   return { text: ack, messages: [], backgrounded: true };
 }
 
@@ -3044,25 +3061,27 @@ var SYSTEM_PROMPT = readAsset(
   "subagents/designExpert/tools/images/enhance-image-prompt.md"
 );
 async function enhanceImagePrompt(params) {
-  const { brief, aspectRatio, transparentBackground, onLog } = params;
-  const orientation = aspectRatio === "1:1" ? "square" : ["16:9", "4:3", "3:2"].includes(aspectRatio) ? "landscape" : "portrait";
+  const { brief, width, height, transparentBackground, onLog } = params;
   const contextParts = [
-    `Aspect ratio: ${aspectRatio} (${orientation})`
+    `Dimensions: ${width}x${height}${width > height ? " (landscape)" : width < height ? " (portrait)" : " (square)"}`
   ];
   if (transparentBackground) {
     contextParts.push(
       "Transparent background: yes \u2014 the background will be removed. Focus on the subject as an isolated element."
     );
   }
-  const message = `<context>
+  const context = `<context>
 ${contextParts.join("\n")}
-</context>
+</context>`;
+  const message = `${SYSTEM_PROMPT}
+
+${context}
 
 <brief>
 ${brief}
 </brief>`;
   const enhanced = await runCli(
-    `mindstudio generate-text --prompt ${JSON.stringify(SYSTEM_PROMPT)} --message ${JSON.stringify(message)} --output-key enhanced --no-meta`,
+    `mindstudio generate-text --message ${JSON.stringify(message)} --output-key content --no-meta`,
     { timeout: 6e4, onLog }
   );
   return enhanced.trim();
@@ -3072,17 +3091,19 @@ ${brief}
 var ANALYZE_PROMPT = "You are reviewing this image for a visual designer sourcing assets for a project. Describe: what the image depicts, the mood and color palette, how the lighting and composition work, any text present in the image, whether there are any issues (artifacts, distortions), and how it could be used in a layout for an app or website. Be concise and practical. Respond only with your analysis as Markdown and absolutely no other text. Do not use emojis - use unicode if you need symbols.";
 async function generateImageAssets(opts) {
   const { prompts, sourceImages, transparentBackground, onLog } = opts;
-  const aspectRatio = opts.aspectRatio || "1:1";
-  const config = {
-    aspect_ratio: aspectRatio,
-    ...sourceImages?.length && { source_images: sourceImages }
-  };
+  const width = opts.width || 2048;
+  const height = opts.height || 2048;
+  const config = { width, height };
+  if (sourceImages?.length) {
+    config.images = sourceImages;
+  }
   const isEdit = !!sourceImages?.length;
   const enhancedPrompts = isEdit ? prompts : await Promise.all(
     prompts.map(
       (brief) => enhanceImagePrompt({
         brief,
-        aspectRatio,
+        width,
+        height,
         transparentBackground,
         onLog
       })
@@ -3093,7 +3114,7 @@ async function generateImageAssets(opts) {
     const step = JSON.stringify({
       prompt: enhancedPrompts[0],
       imageModelOverride: {
-        model: "gemini-3.1-flash-image",
+        model: "seedream-4.5",
         config
       }
     });
@@ -3108,7 +3129,7 @@ async function generateImageAssets(opts) {
       step: {
         prompt,
         imageModelOverride: {
-          model: "gemini-3.1-flash-image",
+          model: "seedream-4.5",
           config
         }
       }
@@ -3161,7 +3182,8 @@ async function generateImageAssets(opts) {
         prompt: prompts[i],
         ...!isEdit && { enhancedPrompt: enhancedPrompts[i] },
         analysis,
-        aspectRatio
+        width,
+        height
       };
     })
   );
@@ -3182,10 +3204,13 @@ var definition6 = {
         },
         description: "One or more image briefs describing what you want. Focus on subject, mood, style, and intended use \u2014 the tool optimizes your brief into a model-ready prompt automatically. Multiple briefs run in parallel."
       },
-      aspectRatio: {
-        type: "string",
-        enum: ["1:1", "16:9", "9:16", "3:4", "4:3", "2:3", "3:2"],
-        description: "Aspect ratio. Default 1:1."
+      width: {
+        type: "number",
+        description: "Image width in pixels. Default 2048. Range: 2048-4096."
+      },
+      height: {
+        type: "number",
+        description: "Image height in pixels. Default 2048. Range: 2048-4096."
       },
       transparentBackground: {
         type: "boolean",
@@ -3198,7 +3223,8 @@ var definition6 = {
 async function execute6(input, onLog) {
   return generateImageAssets({
     prompts: input.prompts,
-    aspectRatio: input.aspectRatio,
+    width: input.width,
+    height: input.height,
     transparentBackground: input.transparentBackground,
     onLog
   });
@@ -3230,10 +3256,13 @@ var definition7 = {
         },
         description: "One or more source/reference image URLs. These are used as the basis for the edit \u2014 the AI will use them as reference for style, subject, or composition."
       },
-      aspectRatio: {
-        type: "string",
-        enum: ["1:1", "16:9", "9:16", "3:4", "4:3", "2:3", "3:2"],
-        description: "Output aspect ratio. Default 1:1."
+      width: {
+        type: "number",
+        description: "Output width in pixels. Default 2048. Range: 2048-4096."
+      },
+      height: {
+        type: "number",
+        description: "Output height in pixels. Default 2048. Range: 2048-4096."
       },
       transparentBackground: {
         type: "boolean",
@@ -3247,7 +3276,8 @@ async function execute7(input, onLog) {
   return generateImageAssets({
     prompts: input.prompts,
     sourceImages: input.sourceImages,
-    aspectRatio: input.aspectRatio,
+    width: input.width,
+    height: input.height,
     transparentBackground: input.transparentBackground,
     onLog
   });
@@ -4728,7 +4758,9 @@ async function runTurn(params) {
         toolRegistry?.register(entry);
         run(tc.input);
         const r = await resultPromise;
-        toolRegistry?.unregister(tc.id);
+        if (!tc.input.background) {
+          toolRegistry?.unregister(tc.id);
+        }
         log7.info("Tool completed", {
           requestId,
           toolCallId: tc.id,

package/dist/index.js

@@ -2163,7 +2163,7 @@ function startStatusWatcher(config) {
     inflight = true;
     try {
       const ctx = getContext();
-      if (!ctx.assistantText && !ctx.lastToolName) {
+      if (!ctx.assistantText && !ctx.lastToolName && !ctx.userMessage) {
         return;
       }
       const res = await fetch(url, {
@@ -2315,7 +2315,8 @@ ${partial}` : "[INTERRUPTED] Agent was interrupted before producing output.",
         getContext: () => ({
           assistantText: getPartialText(contentBlocks),
           lastToolName: currentToolNames || void 0,
-          lastToolResult: lastToolResult || void 0
+          lastToolResult: lastToolResult || void 0,
+          userMessage: task
         }),
         onStatus: (label) => emit2({ type: "status", message: label }),
         signal
@@ -2544,14 +2545,30 @@ Current date/time: ${(/* @__PURE__ */ new Date()).toISOString().replace("T", " "
     return wrapRun();
   }
   log3.info("Sub-agent backgrounded", { requestId, parentToolId, agentName });
+  toolRegistry?.register({
+    id: parentToolId,
+    name: agentName,
+    input: { task },
+    abortController: bgAbort,
+    startedAt: Date.now(),
+    settle: () => {
+    },
+    rerun: () => {
+    },
+    getPartialResult: () => ""
+  });
   const ack = await generateBackgroundAck({
     apiConfig,
     agentName: subAgentId || "agent",
     task
   });
-  wrapRun().then((finalResult) => onBackgroundComplete?.(finalResult)).catch(
-    (err) => onBackgroundComplete?.({ text: `Error: ${err.message}`, messages: [] })
-  );
+  wrapRun().then((finalResult) => {
+    toolRegistry?.unregister(parentToolId);
+    onBackgroundComplete?.(finalResult);
+  }).catch((err) => {
+    toolRegistry?.unregister(parentToolId);
+    onBackgroundComplete?.({ text: `Error: ${err.message}`, messages: [] });
+  });
   return { text: ack, messages: [], backgrounded: true };
 }
 var log3;
@@ -3097,25 +3114,27 @@ var init_screenshot3 = __esm({
 
 // src/subagents/designExpert/tools/images/enhancePrompt.ts
 async function enhanceImagePrompt(params) {
-  const { brief, aspectRatio, transparentBackground, onLog } = params;
-  const orientation = aspectRatio === "1:1" ? "square" : ["16:9", "4:3", "3:2"].includes(aspectRatio) ? "landscape" : "portrait";
+  const { brief, width, height, transparentBackground, onLog } = params;
   const contextParts = [
-    `Aspect ratio: ${aspectRatio} (${orientation})`
+    `Dimensions: ${width}x${height}${width > height ? " (landscape)" : width < height ? " (portrait)" : " (square)"}`
   ];
   if (transparentBackground) {
     contextParts.push(
       "Transparent background: yes \u2014 the background will be removed. Focus on the subject as an isolated element."
     );
   }
-  const message = `<context>
+  const context = `<context>
 ${contextParts.join("\n")}
-</context>
+</context>`;
+  const message = `${SYSTEM_PROMPT}
+
+${context}
 
 <brief>
 ${brief}
 </brief>`;
   const enhanced = await runCli(
-    `mindstudio generate-text --prompt ${JSON.stringify(SYSTEM_PROMPT)} --message ${JSON.stringify(message)} --output-key enhanced --no-meta`,
+    `mindstudio generate-text --message ${JSON.stringify(message)} --output-key content --no-meta`,
     { timeout: 6e4, onLog }
   );
   return enhanced.trim();
@@ -3135,17 +3154,19 @@ var init_enhancePrompt = __esm({
 // src/subagents/designExpert/tools/images/imageGenerator.ts
 async function generateImageAssets(opts) {
   const { prompts, sourceImages, transparentBackground, onLog } = opts;
-  const aspectRatio = opts.aspectRatio || "1:1";
-  const config = {
-    aspect_ratio: aspectRatio,
-    ...sourceImages?.length && { source_images: sourceImages }
-  };
+  const width = opts.width || 2048;
+  const height = opts.height || 2048;
+  const config = { width, height };
+  if (sourceImages?.length) {
+    config.images = sourceImages;
+  }
   const isEdit = !!sourceImages?.length;
   const enhancedPrompts = isEdit ? prompts : await Promise.all(
     prompts.map(
       (brief) => enhanceImagePrompt({
         brief,
-        aspectRatio,
+        width,
+        height,
         transparentBackground,
         onLog
       })
@@ -3156,7 +3177,7 @@ async function generateImageAssets(opts) {
     const step = JSON.stringify({
       prompt: enhancedPrompts[0],
       imageModelOverride: {
-        model: "gemini-3.1-flash-image",
+        model: "seedream-4.5",
         config
       }
     });
@@ -3171,7 +3192,7 @@ async function generateImageAssets(opts) {
       step: {
         prompt,
         imageModelOverride: {
-          model: "gemini-3.1-flash-image",
+          model: "seedream-4.5",
           config
         }
       }
@@ -3224,7 +3245,8 @@ async function generateImageAssets(opts) {
         prompt: prompts[i],
         ...!isEdit && { enhancedPrompt: enhancedPrompts[i] },
         analysis,
-        aspectRatio
+        width,
+        height
       };
     })
   );
@@ -3250,7 +3272,8 @@ __export(generateImages_exports, {
 async function execute6(input, onLog) {
   return generateImageAssets({
     prompts: input.prompts,
-    aspectRatio: input.aspectRatio,
+    width: input.width,
+    height: input.height,
     transparentBackground: input.transparentBackground,
     onLog
   });
@@ -3273,10 +3296,13 @@ var init_generateImages = __esm({
             },
             description: "One or more image briefs describing what you want. Focus on subject, mood, style, and intended use \u2014 the tool optimizes your brief into a model-ready prompt automatically. Multiple briefs run in parallel."
           },
-          aspectRatio: {
-            type: "string",
-            enum: ["1:1", "16:9", "9:16", "3:4", "4:3", "2:3", "3:2"],
-            description: "Aspect ratio. Default 1:1."
+          width: {
+            type: "number",
+            description: "Image width in pixels. Default 2048. Range: 2048-4096."
+          },
+          height: {
+            type: "number",
+            description: "Image height in pixels. Default 2048. Range: 2048-4096."
           },
           transparentBackground: {
             type: "boolean",
@@ -3299,7 +3325,8 @@ async function execute7(input, onLog) {
   return generateImageAssets({
     prompts: input.prompts,
     sourceImages: input.sourceImages,
-    aspectRatio: input.aspectRatio,
+    width: input.width,
+    height: input.height,
     transparentBackground: input.transparentBackground,
     onLog
   });
@@ -3329,10 +3356,13 @@ var init_editImages = __esm({
             },
             description: "One or more source/reference image URLs. These are used as the basis for the edit \u2014 the AI will use them as reference for style, subject, or composition."
           },
-          aspectRatio: {
-            type: "string",
-            enum: ["1:1", "16:9", "9:16", "3:4", "4:3", "2:3", "3:2"],
-            description: "Output aspect ratio. Default 1:1."
+          width: {
+            type: "number",
+            description: "Output width in pixels. Default 2048. Range: 2048-4096."
+          },
+          height: {
+            type: "number",
+            description: "Output height in pixels. Default 2048. Range: 2048-4096."
           },
           transparentBackground: {
             type: "boolean",
@@ -4970,7 +5000,9 @@ async function runTurn(params) {
         toolRegistry?.register(entry);
         run(tc.input);
         const r = await resultPromise;
-        toolRegistry?.unregister(tc.id);
+        if (!tc.input.background) {
+          toolRegistry?.unregister(tc.id);
+        }
         log6.info("Tool completed", {
           requestId,
           toolCallId: tc.id,

package/dist/prompt/compiled/sdk-actions.md

@@ -41,7 +41,6 @@ result.$billingCost;         // cost in credits (if applicable)
 | `generateMusic` | Music from text description | `prompt` | `audioUrl` |
 | `generateLipsync` | Animate face to match audio | `imageUrl`, `audioUrl` | `videoUrl` |
 | `generateAsset` | HTML/PDF/PNG/video output | `prompt` | `assetUrl` |
-| `generateChart` | Chart from data | `data`, `chartType` | `imageUrl` |
 
 ### AI Analysis
 
@@ -133,6 +132,8 @@ const { content } = await agent.generateText({
 });
 ```
 
+Make sure to prioritize new, popular models. MindStudio has a ton of models available - most of them are historical/legacy. Always choose latest generation models from leading providers - Anthropic Claude 4 family, Google Gemini 3, OpenAI GPT 5 to start - the user can adjust later.
+
 ### Batch Execution
 
 Run up to 50 actions in parallel:
@@ -144,3 +145,15 @@ const result = await agent.executeStepBatch([
 ]);
 // result.results[0].output, result.results[1].output
 ```
+
+### AI Generation
+
+MindStudio SDK allows us to build all kinds of amazing AI experiences in apps, including:
+  - Text generation
+  - Image generation (including images with text - AI has gotten good at that now)
+  - Image "remixing" (e.g., user uploads an image, use it as the source image to an image generation model to remix it, or add multiple to generate a collage, etc)
+  - Video generation (including generating video from reference images, start frame images, with audio including voice, etc)
+  - TTS/audio generation
+  - Detailed image and video analysis
+
+Consider the ways in which AI can be incorporated into backend methods to solve problems and be creative.

package/dist/prompt/static/team.md

@@ -59,6 +59,8 @@ When you receive background results:
 
 #### When to Background
 
-Only background the following tasks, unless the user specifically asks you to do work in the background:
+You can only background the following two tasks, unless the user specifically asks you to do work in the background:
 - `productVision` seeding the intiial roadmap after writing the spec for the first time. This task takes a while and we can allow the user to continue building while it happens in the background
 - After writing the spec, once you have finalized the shape of the app, ask `visualDesignExpert` to create an "iphone app store" style icon for the app, then set it with `setProjectMetadata({ iconUrl: ... })`
+
+Do not background any other tasks.

package/dist/subagents/designExpert/tools/images/enhance-image-prompt.md

@@ -32,7 +32,7 @@ These are non-negotiable. Violating them produces bad output.
 
 You'll receive context about the generation parameters. Use them:
 
-- **Aspect ratio**: If the image is landscape (16:9, 4:3, 3:2), compose horizontally. If portrait (9:16, 3:4, 2:3), compose vertically. If square (1:1), center the subject.
+- **Dimensions**: If the image is wide (landscape), compose horizontally. If tall (portrait), compose vertically. If square, center the subject.
 - **Transparent background**: The background will be removed after generation. Don't describe elaborate backgrounds — focus on the subject. Describe it as an isolated element.
 
 ## Photography prompts

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@mindstudio-ai/remy",
-  "version": "0.1.42",
+  "version": "0.1.44",
   "description": "MindStudio coding agent",
   "repository": {
     "type": "git",