@mindstudio-ai/remy 0.1.198 → 0.1.200

package/README.md

@@ -1,8 +1,8 @@
 # Remy
 
-A spec-building and coding agent for MindStudio apps.
+A spec-building and coding agent for building apps.
 
-Remy helps users design, spec, build, and iterate on MindStudio projects. It runs locally in a terminal or as a headless subprocess in the MindStudio sandbox. It has tools for reading/writing specs and code, running shell commands, searching code, prompting users with structured forms, and (in the sandbox) TypeScript language server integration. LLM calls are routed through the MindStudio platform for billing and model routing.
+Remy helps users design, spec, build, and iterate on app projects. It runs locally in a terminal or as a headless subprocess in the sandbox. It has tools for reading/writing specs and code, running shell commands, searching code, prompting users with structured forms, and (in the sandbox) TypeScript language server integration. LLM calls are routed through the Remy platform for billing and model routing.
 
 ## Quick Start
 
@@ -11,7 +11,7 @@ Remy helps users design, spec, build, and iterate on MindStudio projects. It run
 mindstudio login
 
 # Navigate to your project
-cd my-mindstudio-app
+cd my-app
 
 # Run remy
 npx remy
@@ -59,6 +59,7 @@ Tool availability depends on the project's onboarding state, sent by the sandbox
 | `visualDesignExpert` | Visual design expert for fonts, colors, palettes, gradients, layouts, imagery, and icons (sub-agent) |
 | `productVision` | Owns the product roadmap — creates/updates/deletes roadmap items in `src/roadmap/` (sub-agent) |
 | `codeSanityCheck` | Quick readonly sanity check on architecture and package choices before building (sub-agent) |
+| `writeBuildOverview` | Generate/refresh the Build Overview (`src/overview.html`) — a single-page plain-language reference of everything the app contains; you author the copy, the design expert lays it out |
 
 ### Spec Tools
 

package/dist/automatedActions/debugRequest.md

@@ -2,7 +2,7 @@
 trigger: debugRequest
 ---
 
-This is an automated message triggered by the user having clicked "Debug" in the "Request" log detail in MindStudio UI. Find the request in .logs/requests.ndjson by its ID. If there is an error, fix it immediately - otherwise, explain the request at a high-level in non-technical/natural language and see what the user wishes to do with it. Remember, the user can't see this message, so keep that in mind when responding.
+This is an automated message triggered by the user having clicked "Debug" in the "Request" log detail in the Remy UI. Find the request in .logs/requests.ndjson by its ID. If there is an error, fix it immediately - otherwise, explain the request at a high-level in non-technical/natural language and see what the user wishes to do with it. Remember, the user can't see this message, so keep that in mind when responding.
 
 <request_id>
 {{requestId}}

package/dist/automatedActions/feelingLucky.md

@@ -2,7 +2,7 @@
 trigger: feelingLucky
 ---
 
-This is an automated message triggered by the user having clicked "I'm feeling lucky" in the MindStudio UI and selecting a category from a list of app archetypes. They can't see this message. They're likely new to MindStudio and Remy - this might even be their first time using it. They haven't described a specific idea yet and more than likely don't actually have anything specific in mind — they're exploring what they can build and clicked on something that looked interesting. Your job is to ask questions to help them conjure the seed of an idea that is unique, specific, and interesting, and will be compelling to build as an MVP. Don't build something generic - help the user make something unique that allows you to truly show off the power of what you can build. Stay focused on web interfaces for now, unless the user specifically wants a different interface type. When you've aligned on something to build, focus on making it visually beautiful and highly usable from a UX perspective - at this point in the user's experience they're going to be much more "wowed" by something that looks pretty than by something with a ton of backend complexity.
+This is an automated message triggered by the user having clicked "I'm feeling lucky" in the Remy UI and selecting a category from a list of app archetypes. They can't see this message. They're likely new to Remy - this might even be their first time using it. They haven't described a specific idea yet and more than likely don't actually have anything specific in mind — they're exploring what they can build and clicked on something that looked interesting. Your job is to ask questions to help them conjure the seed of an idea that is unique, specific, and interesting, and will be compelling to build as an MVP. Don't build something generic - help the user make something unique that allows you to truly show off the power of what you can build. Stay focused on web interfaces for now, unless the user specifically wants a different interface type. When you've aligned on something to build, focus on making it visually beautiful and highly usable from a UX perspective - at this point in the user's experience they're going to be much more "wowed" by something that looks pretty than by something with a ton of backend complexity.
 
 <selected_category>
 {{title}}

package/dist/automatedActions/postBuildPolish.md

@@ -13,6 +13,7 @@ The visual assets — photography, generated images, brand colors, typography
 
 ## Finalizing
 When everything is working and polished:
-1. Use `productVision` to mark the MVP roadmap item as done.
-2. Call `compactConversation` to summarize the build session and free up context for the next phase of work.
-3. Call `setProjectOnboardingState({ state: "buildComplete" })`. This triggers the reveal experience on the frontend, where the user sees the pitch deck and their finished app for the first time.
+1. Write the Build Overview. Author the full copy of everything this build actually produced — the data stores and what each holds, the backend operations, the interfaces, any access and role boundaries, background jobs, and seeded scenarios, including the parts the user can't see — in plain, calm, present-tense language, with real names and exact counts. Then call `writeBuildOverview` to render it. Do this now, before compacting, while the build is still fresh in context.
+2. Use `productVision` to mark the MVP roadmap item as done.
+3. Call `compactConversation` to summarize the build session and free up context for the next phase of work.
+4. Call `setProjectOnboardingState({ state: "buildComplete" })`. This triggers the reveal experience on the frontend, where the user sees their finished app and its Build Overview for the first time.

package/dist/brandExtraction/extract.md

@@ -1,4 +1,4 @@
-You extract a structured `AppBrand` JSON object from the spec files of a MindStudio app project.
+You extract a structured `AppBrand` JSON object from the spec files of an app project.
 
 Your output is read by a frontend renderer that uses the brand to style internal documents (implementation plans, sync plans, publish plans) with a "letterhead" treatment — a wordmark, accent color, paper-tone background, and branded fonts. Every field is optional. The renderer falls back to generic styling when a field is missing or invalid.
 

package/dist/compaction/conversation.md

@@ -1,4 +1,4 @@
-You are summarizing a conversation between a user and an AI coding agent called Remy that builds MindStudio apps. Your summary will replace the conversation history — the agent will only see your summary plus any new messages, so it must capture everything needed to continue working.
+You are summarizing a conversation between a user and an AI coding agent called Remy that builds apps. Your summary will replace the conversation history — the agent will only see your summary plus any new messages, so it must capture everything needed to continue working.
 
 Write a tight, factual summary in bullet points. Cover what matters for picking the work back up:
 - What the user is trying to do, and any key decisions, constraints, or preferences they've stated.

package/dist/headless.js

@@ -4232,7 +4232,15 @@ var SYSTEM_PROMPT = readAsset(
   "subagents/designExpert/tools/images/enhance-image-prompt.md"
 );
 async function enhanceImagePrompt(params) {
-  const { brief, width, height, transparentBackground, onLog, model } = params;
+  const {
+    brief,
+    width,
+    height,
+    transparentBackground,
+    hasReferenceImage,
+    onLog,
+    model
+  } = params;
   const contextParts = [
     `Dimensions: ${width}x${height}${width > height ? " (landscape)" : width < height ? " (portrait)" : " (square)"}`
   ];
@@ -4241,6 +4249,11 @@ async function enhanceImagePrompt(params) {
       "Transparent background: yes \u2014 the background will be removed. Focus on the subject as an isolated element."
     );
   }
+  if (hasReferenceImage) {
+    contextParts.push(
+      "Reference image: yes \u2014 a reference image is provided to the generation model alongside this prompt to guide style, subject, or composition. Complement it; don't re-describe what it already carries."
+    );
+  }
   const context = `<context>
 ${contextParts.join("\n")}
 </context>`;
@@ -4271,6 +4284,7 @@ async function generateImageAssets(opts) {
     prompts,
     sourceImages,
     transparentBackground,
+    enhancePrompts,
     onLog,
     imageGenerationModel: genModel,
     imageAnalysisModel,
@@ -4289,19 +4303,20 @@ async function generateImageAssets(opts) {
     config.source_image = firstImage;
     config.source = firstImage;
   }
-  const isEdit = !!sourceImages?.length;
-  const enhancedPrompts = isEdit ? prompts : await Promise.all(
+  const hasReference = !!sourceImages?.length;
+  const enhancedPrompts = enhancePrompts ? await Promise.all(
     prompts.map(
       (brief) => enhanceImagePrompt({
         brief,
         width,
         height,
         transparentBackground,
+        hasReferenceImage: hasReference,
         onLog,
         model: imagePromptEnhancerModel
       })
     )
-  );
+  ) : prompts;
   let imageUrls;
   if (enhancedPrompts.length === 1) {
     const step = JSON.stringify({
@@ -4371,7 +4386,7 @@ async function generateImageAssets(opts) {
       if (url.startsWith("Error")) {
         return {
           prompt: prompts[i],
-          ...!isEdit && { enhancedPrompt: enhancedPrompts[i] },
+          ...enhancePrompts && { enhancedPrompt: enhancedPrompts[i] },
           error: url
         };
       }
@@ -4384,7 +4399,7 @@ async function generateImageAssets(opts) {
       return {
         url,
         prompt: prompts[i],
-        ...!isEdit && { enhancedPrompt: enhancedPrompts[i] },
+        ...enhancePrompts && { enhancedPrompt: enhancedPrompts[i] },
         analysis,
         width,
         height
@@ -4409,6 +4424,10 @@ 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."
       },
+      referenceImage: {
+        type: "string",
+        description: "Optional URL of a single reference image to guide the generation \u2014 for style, subject, character consistency, or composition. Your prompt still describes the desired result; the reference conditions it. Applies to every prompt in the batch."
+      },
       width: {
         type: "number",
         description: "Image width in pixels. Default 2048. Range: 2048-4096."
@@ -4431,6 +4450,8 @@ async function execute6(input, onLog, context) {
     width: input.width,
     height: input.height,
     transparentBackground: input.transparentBackground,
+    sourceImages: input.referenceImage ? [input.referenceImage] : void 0,
+    enhancePrompts: true,
     onLog,
     imageGenerationModel: resolveModel(
       "imageGeneration",
@@ -4500,6 +4521,7 @@ async function execute7(input, onLog, context) {
     width: input.width,
     height: input.height,
     transparentBackground: input.transparentBackground,
+    enhancePrompts: false,
     onLog,
     imageGenerationModel: resolveModel(
       "imageGeneration",
@@ -4658,9 +4680,9 @@ ${parts.join("\n\n")}
 }
 function loadPlatformBrief() {
   return `<platform_brief>
-## What is a MindStudio app?
+## What is a Remy app?
 
-A MindStudio app is a managed full-stack TypeScript project with three layers: a spec (natural language in src/), a backend contract (methods, tables, roles in dist/), and one or more interfaces (web, API, bots, cron, etc.). The spec is the source of truth; code is derived from it.
+A Remy app is a managed full-stack TypeScript project with three layers: a spec (natural language in src/), a backend contract (methods, tables, roles in dist/), and one or more interfaces (web, API, bots, cron, etc.). The spec is the source of truth; code is derived from it.
 
 This is a capable, stable platform used in production by 100k+ users. Build with confidence \u2014 you're building production-grade apps, not fragile prototypes.
 
@@ -4705,7 +4727,7 @@ TypeScript running in a sandboxed environment. Any npm package can be installed.
 
 The first-party SDK (@mindstudio-ai/agent) provides access to 200+ AI models (OpenAI, Anthropic, Google, Meta, Mistral, and more) and 1000+ integrations (email, SMS, Slack, HubSpot, Google Workspace, web scraping, image/video generation, media processing, and much more) with zero configuration \u2014 credentials are handled automatically in the execution environment. No API keys needed. This SDK is robust and battle-tested in production.
 
-## What MindStudio apps are NOT good for
+## What Remy apps are NOT good for
 
 - Native mobile apps (iOS/Android). Mobile-responsive web apps are fine.
 - Real-time multiplayer with persistent connections (no WebSocket support). Turn-based or async patterns work.
@@ -5369,6 +5391,94 @@ var scrapeWebUrlTool = {
   }
 };
 
+// src/tools/spec/writeBuildOverview.ts
+import fs20 from "fs";
+var OVERVIEW_FILE = "src/overview.html";
+var DESIGN_BRIEF = `We are building the Build Overview for this app \u2014 the home page of its Spec tab. It is a calm, dense, one-page reference of everything the app actually contains, including the parts the user can't see. It renders flush inside the Spec tab's content panel (the IDE supplies the surrounding nav).
+
+Take the plain-language copy in <overview_copy> and lay it out and skin it into a single, beautiful, self-contained HTML document in the app's own brand. If <current_overview> is non-empty, use it as your starting point and preserve its established skin, updating only what the copy changed.
+
+### The single hard rule
+Preserve every fact in <overview_copy> exactly \u2014 every number, name, label, and claim. Do not add, drop, soften, reword, or invent any fact. You own layout, typography, and visual design only; the substance and the words are fixed. A single wrong number breaks this document's entire purpose.
+
+### What it is (and is not)
+- A typeset reference dossier: composed, dense, a little cool \u2014 substantial at a glance, then readable. Density communicates substance; sparse and airy reads as "not much here."
+- Brand-skinned: pull the palette, type system, and a single accent from the app's spec. The accent leads labels and figures; it is a mark, not a fill.
+- NOT a slide deck \u2014 no theatrics, no building to a climax, no persuasion.
+- NOT a docs site \u2014 no raised or shadowed cards, no hover-lift, no clickable-feeling surfaces, no sticky table of contents.
+- NOT a sparse memo \u2014 a restrained single-column page under-sells the work.
+- No header chrome and no footer: open straight on the app's logo and name so it sits flush with the IDE.
+
+### Constraints
+- A single self-contained HTML file. Fonts may load from a CDN; everything else (CSS, the logo SVG) is inline.
+- Responsive: fills the embedded panel width and collapses gracefully at narrow widths.
+
+Respond only with the complete HTML file and absolutely no other text. Your response will be written directly to src/overview.html.`;
+var OVERVIEW_SHELL = `<!DOCTYPE html>
+<html lang="en">
+<head>
+<meta charset="utf-8">
+<!-- This renders inside an iframe in the IDE. Keep it a simple, self-contained
+     static document \u2014 no scroll/zoom/accessibility scaffolding beyond this head.
+     The user-scalable=no viewport is intentional; leave it. -->
+<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no">
+<title>Build Overview</title>
+<!-- SKIN: link the app's fonts here (CDN \u2014 e.g. Fontshare / Google Fonts). -->
+<style>
+*, *::before, *::after { margin: 0; padding: 0; box-sizing: border-box; }
+</style>
+</head>
+<body>
+<!-- Compose the Build Overview here. Layout, sections, type, and styling are
+     yours \u2014 keep everything inline and self-contained, brand-skinned from the
+     app's spec. -->
+</body>
+</html>`;
+var buildOverviewTool = {
+  clearable: false,
+  definition: {
+    name: "writeBuildOverview",
+    description: "Generate or refresh the Build Overview \u2014 the project's home page in the Spec tab: a single-page, plain-language reference of everything the app actually contains, including the parts the user can't see (data stores, backend operations, access and roles, background jobs, seeded scenarios, the design system). You author the full copy: read the manifest and spec and state, plainly and exactly, what genuinely exists \u2014 real names and accurate counts \u2014 in calm, declarative, present-tense outcome language, with no persuasion or hype. Describe only what exists. Pass the complete copy as `content`; the design expert lays it out and skins it to the app's brand without altering any of your facts. Generate it at the end of a build and refresh it after meaningful work.",
+    inputSchema: {
+      type: "object",
+      properties: {
+        content: {
+          type: "string",
+          description: "The full Build Overview copy you authored: everything the app contains, in plain present-tense outcome language, with real names and exact counts. The design expert lays this out and skins it to the brand without changing any fact."
+        }
+      },
+      required: ["content"]
+    }
+  },
+  async execute(input, context) {
+    if (!context) {
+      return "Error: writeBuildOverview requires execution context for design expert delegation";
+    }
+    const content = (input.content ?? "").trim();
+    if (!content) {
+      return "Error: writeBuildOverview requires non-empty `content` (the overview copy).";
+    }
+    try {
+      const existing = fs20.existsSync(OVERVIEW_FILE) ? fs20.readFileSync(OVERVIEW_FILE, "utf-8").trim() : "";
+      const currentOverview = existing || OVERVIEW_SHELL;
+      const task = `<overview_copy>${content}</overview_copy>
+
+<current_overview>${currentOverview}</current_overview>
+
+${DESIGN_BRIEF}`;
+      const result = await designExpertTool.execute({ task }, context);
+      const htmlMatch = result.match(
+        /```(?:html|wireframe)\n([\s\S]*?)```/
+      );
+      const html = htmlMatch ? htmlMatch[1].trim() : result;
+      fs20.writeFileSync(OVERVIEW_FILE, html, "utf-8");
+      return "Build overview written successfully to src/overview.html.";
+    } catch (err) {
+      return `Error generating build overview: ${err.message}`;
+    }
+  }
+};
+
 // src/tools/index.ts
 function deriveContext(parent, toolCallId) {
   return { ...parent, toolCallId };
@@ -5385,6 +5495,7 @@ var ALL_TOOLS = [
   designExpertTool,
   productVisionTool,
   codeSanityCheckTool,
+  buildOverviewTool,
   compactConversationTool,
   // Post-onboarding
   presentPublishPlanTool,
@@ -5767,7 +5878,7 @@ function triggerCompaction(state, apiConfig, opts = {}) {
 }
 
 // src/brandExtraction/index.ts
-import fs20 from "fs";
+import fs21 from "fs";
 import path10 from "path";
 import { createHash } from "crypto";
 var log9 = createLogger("brandExtraction");
@@ -5816,7 +5927,7 @@ function sha256(input) {
 }
 function readSafe(filePath) {
   try {
-    return fs20.readFileSync(filePath, "utf-8");
+    return fs21.readFileSync(filePath, "utf-8");
   } catch {
     return "";
   }
@@ -5824,7 +5935,7 @@ function readSafe(filePath) {
 function walkMdFiles3(dir) {
   const results = [];
   try {
-    const entries = fs20.readdirSync(dir, { withFileTypes: true });
+    const entries = fs21.readdirSync(dir, { withFileTypes: true });
     for (const entry of entries) {
       const full = path10.join(dir, entry.name);
       if (entry.isDirectory()) {
@@ -5839,7 +5950,7 @@ function walkMdFiles3(dir) {
 }
 function parseFrontmatter3(filePath) {
   try {
-    const content = fs20.readFileSync(filePath, "utf-8");
+    const content = fs21.readFileSync(filePath, "utf-8");
     const match = content.match(/^---\n([\s\S]*?)\n---/);
     if (!match) {
       return { type: "" };
@@ -6018,14 +6129,14 @@ function pickFont(raw) {
 }
 function persistBrand(brand, inputHash) {
   const tmp = `${BRAND_FILE}.tmp`;
-  fs20.writeFileSync(tmp, JSON.stringify(brand, null, 2), "utf-8");
-  fs20.renameSync(tmp, BRAND_FILE);
+  fs21.writeFileSync(tmp, JSON.stringify(brand, null, 2), "utf-8");
+  fs21.renameSync(tmp, BRAND_FILE);
   const cache = { inputHash, generatedAt: Date.now() };
-  fs20.writeFileSync(CACHE_FILE, JSON.stringify(cache, null, 2), "utf-8");
+  fs21.writeFileSync(CACHE_FILE, JSON.stringify(cache, null, 2), "utf-8");
 }
 function readCache() {
   try {
-    const raw = fs20.readFileSync(CACHE_FILE, "utf-8");
+    const raw = fs21.readFileSync(CACHE_FILE, "utf-8");
     const parsed = JSON.parse(raw);
     if (parsed && typeof parsed.inputHash === "string" && typeof parsed.generatedAt === "number") {
       return parsed;
@@ -6058,14 +6169,14 @@ function triggerBrandExtraction(apiConfig, model) {
 }
 
 // src/session.ts
-import fs21 from "fs";
+import fs22 from "fs";
 import path11 from "path";
 var log11 = createLogger("session");
 var SESSION_FILE = ".remy-session.json";
 var ARCHIVE_DIR = ".logs/sessions";
 function loadSession(state) {
   try {
-    const raw = fs21.readFileSync(SESSION_FILE, "utf-8");
+    const raw = fs22.readFileSync(SESSION_FILE, "utf-8");
     const data = JSON.parse(raw);
     if (data.models && typeof data.models === "object") {
       state.models = data.models;
@@ -6124,7 +6235,7 @@ function saveSession(state) {
     if (state.models && Object.keys(state.models).length > 0) {
       payload.models = state.models;
     }
-    fs21.writeFileSync(SESSION_FILE, JSON.stringify(payload, null, 2), "utf-8");
+    fs22.writeFileSync(SESSION_FILE, JSON.stringify(payload, null, 2), "utf-8");
     log11.info("Session saved", { messageCount: state.messages.length });
   } catch (err) {
     log11.warn("Session save failed", { error: err.message });
@@ -6133,11 +6244,11 @@ function saveSession(state) {
 function clearSession(state) {
   state.messages = [];
   try {
-    if (fs21.existsSync(SESSION_FILE)) {
-      fs21.mkdirSync(ARCHIVE_DIR, { recursive: true });
+    if (fs22.existsSync(SESSION_FILE)) {
+      fs22.mkdirSync(ARCHIVE_DIR, { recursive: true });
       const ts = (/* @__PURE__ */ new Date()).toISOString().replace(/[:.]/g, "-");
       const dest = path11.join(ARCHIVE_DIR, `cleared-${ts}.json`);
-      fs21.renameSync(SESSION_FILE, dest);
+      fs22.renameSync(SESSION_FILE, dest);
       log11.info("Session archived on clear", { dest });
     }
   } catch (err) {
@@ -6145,7 +6256,7 @@ function clearSession(state) {
       error: err.message
     });
     try {
-      fs21.unlinkSync(SESSION_FILE);
+      fs22.unlinkSync(SESSION_FILE);
     } catch {
     }
   }

package/dist/index.js

@@ -4957,7 +4957,15 @@ var init_screenshot3 = __esm({
 
 // src/subagents/designExpert/tools/images/enhancePrompt.ts
 async function enhanceImagePrompt(params) {
-  const { brief, width, height, transparentBackground, onLog, model } = params;
+  const {
+    brief,
+    width,
+    height,
+    transparentBackground,
+    hasReferenceImage,
+    onLog,
+    model
+  } = params;
   const contextParts = [
     `Dimensions: ${width}x${height}${width > height ? " (landscape)" : width < height ? " (portrait)" : " (square)"}`
   ];
@@ -4966,6 +4974,11 @@ async function enhanceImagePrompt(params) {
       "Transparent background: yes \u2014 the background will be removed. Focus on the subject as an isolated element."
     );
   }
+  if (hasReferenceImage) {
+    contextParts.push(
+      "Reference image: yes \u2014 a reference image is provided to the generation model alongside this prompt to guide style, subject, or composition. Complement it; don't re-describe what it already carries."
+    );
+  }
   const context = `<context>
 ${contextParts.join("\n")}
 </context>`;
@@ -5006,6 +5019,7 @@ async function generateImageAssets(opts) {
     prompts,
     sourceImages,
     transparentBackground,
+    enhancePrompts,
     onLog,
     imageGenerationModel: genModel,
     imageAnalysisModel,
@@ -5024,19 +5038,20 @@ async function generateImageAssets(opts) {
     config.source_image = firstImage;
     config.source = firstImage;
   }
-  const isEdit = !!sourceImages?.length;
-  const enhancedPrompts = isEdit ? prompts : await Promise.all(
+  const hasReference = !!sourceImages?.length;
+  const enhancedPrompts = enhancePrompts ? await Promise.all(
     prompts.map(
       (brief) => enhanceImagePrompt({
         brief,
         width,
         height,
         transparentBackground,
+        hasReferenceImage: hasReference,
         onLog,
         model: imagePromptEnhancerModel
       })
     )
-  );
+  ) : prompts;
   let imageUrls;
   if (enhancedPrompts.length === 1) {
     const step = JSON.stringify({
@@ -5106,7 +5121,7 @@ async function generateImageAssets(opts) {
       if (url.startsWith("Error")) {
         return {
           prompt: prompts[i],
-          ...!isEdit && { enhancedPrompt: enhancedPrompts[i] },
+          ...enhancePrompts && { enhancedPrompt: enhancedPrompts[i] },
           error: url
         };
       }
@@ -5119,7 +5134,7 @@ async function generateImageAssets(opts) {
       return {
         url,
         prompt: prompts[i],
-        ...!isEdit && { enhancedPrompt: enhancedPrompts[i] },
+        ...enhancePrompts && { enhancedPrompt: enhancedPrompts[i] },
         analysis,
         width,
         height
@@ -5151,6 +5166,8 @@ async function execute6(input, onLog, context) {
     width: input.width,
     height: input.height,
     transparentBackground: input.transparentBackground,
+    sourceImages: input.referenceImage ? [input.referenceImage] : void 0,
+    enhancePrompts: true,
     onLog,
     imageGenerationModel: resolveModel(
       "imageGeneration",
@@ -5189,6 +5206,10 @@ 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."
           },
+          referenceImage: {
+            type: "string",
+            description: "Optional URL of a single reference image to guide the generation \u2014 for style, subject, character consistency, or composition. Your prompt still describes the desired result; the reference conditions it. Applies to every prompt in the batch."
+          },
           width: {
             type: "number",
             description: "Image width in pixels. Default 2048. Range: 2048-4096."
@@ -5221,6 +5242,7 @@ async function execute7(input, onLog, context) {
     width: input.width,
     height: input.height,
     transparentBackground: input.transparentBackground,
+    enhancePrompts: false,
     onLog,
     imageGenerationModel: resolveModel(
       "imageGeneration",
@@ -5439,9 +5461,9 @@ ${parts.join("\n\n")}
 }
 function loadPlatformBrief() {
   return `<platform_brief>
-## What is a MindStudio app?
+## What is a Remy app?
 
-A MindStudio app is a managed full-stack TypeScript project with three layers: a spec (natural language in src/), a backend contract (methods, tables, roles in dist/), and one or more interfaces (web, API, bots, cron, etc.). The spec is the source of truth; code is derived from it.
+A Remy app is a managed full-stack TypeScript project with three layers: a spec (natural language in src/), a backend contract (methods, tables, roles in dist/), and one or more interfaces (web, API, bots, cron, etc.). The spec is the source of truth; code is derived from it.
 
 This is a capable, stable platform used in production by 100k+ users. Build with confidence \u2014 you're building production-grade apps, not fragile prototypes.
 
@@ -5486,7 +5508,7 @@ TypeScript running in a sandboxed environment. Any npm package can be installed.
 
 The first-party SDK (@mindstudio-ai/agent) provides access to 200+ AI models (OpenAI, Anthropic, Google, Meta, Mistral, and more) and 1000+ integrations (email, SMS, Slack, HubSpot, Google Workspace, web scraping, image/video generation, media processing, and much more) with zero configuration \u2014 credentials are handled automatically in the execution environment. No API keys needed. This SDK is robust and battle-tested in production.
 
-## What MindStudio apps are NOT good for
+## What Remy apps are NOT good for
 
 - Native mobile apps (iOS/Android). Mobile-responsive web apps are fine.
 - Real-time multiplayer with persistent connections (no WebSocket support). Turn-based or async patterns work.
@@ -6281,6 +6303,101 @@ var init_scrapeWebUrl2 = __esm({
   }
 });
 
+// src/tools/spec/writeBuildOverview.ts
+import fs19 from "fs";
+var OVERVIEW_FILE, DESIGN_BRIEF, OVERVIEW_SHELL, buildOverviewTool;
+var init_writeBuildOverview = __esm({
+  "src/tools/spec/writeBuildOverview.ts"() {
+    "use strict";
+    init_designExpert();
+    OVERVIEW_FILE = "src/overview.html";
+    DESIGN_BRIEF = `We are building the Build Overview for this app \u2014 the home page of its Spec tab. It is a calm, dense, one-page reference of everything the app actually contains, including the parts the user can't see. It renders flush inside the Spec tab's content panel (the IDE supplies the surrounding nav).
+
+Take the plain-language copy in <overview_copy> and lay it out and skin it into a single, beautiful, self-contained HTML document in the app's own brand. If <current_overview> is non-empty, use it as your starting point and preserve its established skin, updating only what the copy changed.
+
+### The single hard rule
+Preserve every fact in <overview_copy> exactly \u2014 every number, name, label, and claim. Do not add, drop, soften, reword, or invent any fact. You own layout, typography, and visual design only; the substance and the words are fixed. A single wrong number breaks this document's entire purpose.
+
+### What it is (and is not)
+- A typeset reference dossier: composed, dense, a little cool \u2014 substantial at a glance, then readable. Density communicates substance; sparse and airy reads as "not much here."
+- Brand-skinned: pull the palette, type system, and a single accent from the app's spec. The accent leads labels and figures; it is a mark, not a fill.
+- NOT a slide deck \u2014 no theatrics, no building to a climax, no persuasion.
+- NOT a docs site \u2014 no raised or shadowed cards, no hover-lift, no clickable-feeling surfaces, no sticky table of contents.
+- NOT a sparse memo \u2014 a restrained single-column page under-sells the work.
+- No header chrome and no footer: open straight on the app's logo and name so it sits flush with the IDE.
+
+### Constraints
+- A single self-contained HTML file. Fonts may load from a CDN; everything else (CSS, the logo SVG) is inline.
+- Responsive: fills the embedded panel width and collapses gracefully at narrow widths.
+
+Respond only with the complete HTML file and absolutely no other text. Your response will be written directly to src/overview.html.`;
+    OVERVIEW_SHELL = `<!DOCTYPE html>
+<html lang="en">
+<head>
+<meta charset="utf-8">
+<!-- This renders inside an iframe in the IDE. Keep it a simple, self-contained
+     static document \u2014 no scroll/zoom/accessibility scaffolding beyond this head.
+     The user-scalable=no viewport is intentional; leave it. -->
+<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no">
+<title>Build Overview</title>
+<!-- SKIN: link the app's fonts here (CDN \u2014 e.g. Fontshare / Google Fonts). -->
+<style>
+*, *::before, *::after { margin: 0; padding: 0; box-sizing: border-box; }
+</style>
+</head>
+<body>
+<!-- Compose the Build Overview here. Layout, sections, type, and styling are
+     yours \u2014 keep everything inline and self-contained, brand-skinned from the
+     app's spec. -->
+</body>
+</html>`;
+    buildOverviewTool = {
+      clearable: false,
+      definition: {
+        name: "writeBuildOverview",
+        description: "Generate or refresh the Build Overview \u2014 the project's home page in the Spec tab: a single-page, plain-language reference of everything the app actually contains, including the parts the user can't see (data stores, backend operations, access and roles, background jobs, seeded scenarios, the design system). You author the full copy: read the manifest and spec and state, plainly and exactly, what genuinely exists \u2014 real names and accurate counts \u2014 in calm, declarative, present-tense outcome language, with no persuasion or hype. Describe only what exists. Pass the complete copy as `content`; the design expert lays it out and skins it to the app's brand without altering any of your facts. Generate it at the end of a build and refresh it after meaningful work.",
+        inputSchema: {
+          type: "object",
+          properties: {
+            content: {
+              type: "string",
+              description: "The full Build Overview copy you authored: everything the app contains, in plain present-tense outcome language, with real names and exact counts. The design expert lays this out and skins it to the brand without changing any fact."
+            }
+          },
+          required: ["content"]
+        }
+      },
+      async execute(input, context) {
+        if (!context) {
+          return "Error: writeBuildOverview requires execution context for design expert delegation";
+        }
+        const content = (input.content ?? "").trim();
+        if (!content) {
+          return "Error: writeBuildOverview requires non-empty `content` (the overview copy).";
+        }
+        try {
+          const existing = fs19.existsSync(OVERVIEW_FILE) ? fs19.readFileSync(OVERVIEW_FILE, "utf-8").trim() : "";
+          const currentOverview = existing || OVERVIEW_SHELL;
+          const task = `<overview_copy>${content}</overview_copy>
+
+<current_overview>${currentOverview}</current_overview>
+
+${DESIGN_BRIEF}`;
+          const result = await designExpertTool.execute({ task }, context);
+          const htmlMatch = result.match(
+            /```(?:html|wireframe)\n([\s\S]*?)```/
+          );
+          const html = htmlMatch ? htmlMatch[1].trim() : result;
+          fs19.writeFileSync(OVERVIEW_FILE, html, "utf-8");
+          return "Build overview written successfully to src/overview.html.";
+        } catch (err) {
+          return `Error generating build overview: ${err.message}`;
+        }
+      }
+    };
+  }
+});
+
 // src/tools/index.ts
 function deriveContext(parent, toolCallId) {
   return { ...parent, toolCallId };
@@ -6335,6 +6452,7 @@ var init_tools6 = __esm({
     init_productVision();
     init_codeSanityCheck();
     init_scrapeWebUrl2();
+    init_writeBuildOverview();
     ALL_TOOLS = [
       // Common
       setProjectOnboardingStateTool,
@@ -6347,6 +6465,7 @@ var init_tools6 = __esm({
       designExpertTool,
       productVisionTool,
       codeSanityCheckTool,
+      buildOverviewTool,
       compactConversationTool,
       // Post-onboarding
       presentPublishPlanTool,
@@ -6389,11 +6508,11 @@ var init_tools6 = __esm({
 });
 
 // src/session.ts
-import fs19 from "fs";
+import fs20 from "fs";
 import path9 from "path";
 function loadSession(state) {
   try {
-    const raw = fs19.readFileSync(SESSION_FILE, "utf-8");
+    const raw = fs20.readFileSync(SESSION_FILE, "utf-8");
     const data = JSON.parse(raw);
     if (data.models && typeof data.models === "object") {
       state.models = data.models;
@@ -6452,7 +6571,7 @@ function saveSession(state) {
     if (state.models && Object.keys(state.models).length > 0) {
       payload.models = state.models;
     }
-    fs19.writeFileSync(SESSION_FILE, JSON.stringify(payload, null, 2), "utf-8");
+    fs20.writeFileSync(SESSION_FILE, JSON.stringify(payload, null, 2), "utf-8");
     log8.info("Session saved", { messageCount: state.messages.length });
   } catch (err) {
     log8.warn("Session save failed", { error: err.message });
@@ -6461,11 +6580,11 @@ function saveSession(state) {
 function clearSession(state) {
   state.messages = [];
   try {
-    if (fs19.existsSync(SESSION_FILE)) {
-      fs19.mkdirSync(ARCHIVE_DIR, { recursive: true });
+    if (fs20.existsSync(SESSION_FILE)) {
+      fs20.mkdirSync(ARCHIVE_DIR, { recursive: true });
       const ts = (/* @__PURE__ */ new Date()).toISOString().replace(/[:.]/g, "-");
       const dest = path9.join(ARCHIVE_DIR, `cleared-${ts}.json`);
-      fs19.renameSync(SESSION_FILE, dest);
+      fs20.renameSync(SESSION_FILE, dest);
       log8.info("Session archived on clear", { dest });
     }
   } catch (err) {
@@ -6473,7 +6592,7 @@ function clearSession(state) {
       error: err.message
     });
     try {
-      fs19.unlinkSync(SESSION_FILE);
+      fs20.unlinkSync(SESSION_FILE);
     } catch {
     }
   }
@@ -6695,7 +6814,7 @@ var init_errors = __esm({
 });
 
 // src/brandExtraction/index.ts
-import fs20 from "fs";
+import fs21 from "fs";
 import path10 from "path";
 import { createHash } from "crypto";
 async function runExtraction(apiConfig, model) {
@@ -6740,7 +6859,7 @@ function sha256(input) {
 }
 function readSafe(filePath) {
   try {
-    return fs20.readFileSync(filePath, "utf-8");
+    return fs21.readFileSync(filePath, "utf-8");
   } catch {
     return "";
   }
@@ -6748,7 +6867,7 @@ function readSafe(filePath) {
 function walkMdFiles3(dir) {
   const results = [];
   try {
-    const entries = fs20.readdirSync(dir, { withFileTypes: true });
+    const entries = fs21.readdirSync(dir, { withFileTypes: true });
     for (const entry of entries) {
       const full = path10.join(dir, entry.name);
       if (entry.isDirectory()) {
@@ -6763,7 +6882,7 @@ function walkMdFiles3(dir) {
 }
 function parseFrontmatter3(filePath) {
   try {
-    const content = fs20.readFileSync(filePath, "utf-8");
+    const content = fs21.readFileSync(filePath, "utf-8");
     const match = content.match(/^---\n([\s\S]*?)\n---/);
     if (!match) {
       return { type: "" };
@@ -6942,14 +7061,14 @@ function pickFont(raw) {
 }
 function persistBrand(brand, inputHash) {
   const tmp = `${BRAND_FILE}.tmp`;
-  fs20.writeFileSync(tmp, JSON.stringify(brand, null, 2), "utf-8");
-  fs20.renameSync(tmp, BRAND_FILE);
+  fs21.writeFileSync(tmp, JSON.stringify(brand, null, 2), "utf-8");
+  fs21.renameSync(tmp, BRAND_FILE);
   const cache = { inputHash, generatedAt: Date.now() };
-  fs20.writeFileSync(CACHE_FILE, JSON.stringify(cache, null, 2), "utf-8");
+  fs21.writeFileSync(CACHE_FILE, JSON.stringify(cache, null, 2), "utf-8");
 }
 function readCache() {
   try {
-    const raw = fs20.readFileSync(CACHE_FILE, "utf-8");
+    const raw = fs21.readFileSync(CACHE_FILE, "utf-8");
     const parsed = JSON.parse(raw);
     if (parsed && typeof parsed.inputHash === "string" && typeof parsed.generatedAt === "number") {
       return parsed;
@@ -7616,12 +7735,12 @@ var init_agent = __esm({
 });
 
 // src/config.ts
-import fs21 from "fs";
+import fs22 from "fs";
 import path11 from "path";
 import os from "os";
 function loadConfigFile() {
   try {
-    const raw = fs21.readFileSync(CONFIG_PATH, "utf-8");
+    const raw = fs22.readFileSync(CONFIG_PATH, "utf-8");
     log12.debug("Loaded config file", { path: CONFIG_PATH });
     return JSON.parse(raw);
   } catch (err) {
@@ -8869,7 +8988,7 @@ var init_headless = __esm({
 // src/index.tsx
 import { render } from "ink";
 import os2 from "os";
-import fs22 from "fs";
+import fs23 from "fs";
 import path12 from "path";
 
 // src/tui/App.tsx
@@ -9151,7 +9270,7 @@ Error: ${err.message}`,
   return /* @__PURE__ */ jsxs4(Box4, { flexDirection: "column", gap: 1, children: [
     /* @__PURE__ */ jsxs4(Text5, { bold: true, color: "magenta", children: [
       "Remy ",
-      /* @__PURE__ */ jsx5(Text5, { dimColor: true, children: "v0.1.0 \u2014 MindStudio coding agent" })
+      /* @__PURE__ */ jsx5(Text5, { dimColor: true, children: "v0.1.0 \u2014 AI coding agent" })
     ] }),
     sessionRestored && /* @__PURE__ */ jsxs4(Text5, { dimColor: true, children: [
       "Session restored (",
@@ -9188,7 +9307,7 @@ for (let i = 0; i < args.length; i++) {
 var startupLog = createLogger("startup");
 function printDebugInfo(config) {
   const pkg = JSON.parse(
-    fs22.readFileSync(
+    fs23.readFileSync(
       path12.join(import.meta.dirname, "..", "package.json"),
       "utf-8"
     )

package/dist/prompt/.notes.md

@@ -69,7 +69,7 @@ From research into v0, Lovable, Bolt, and Anthropic's `<frontend_aesthetics>` co
 
 ## SDK Usage
 
-The canonical import pattern for MindStudio app methods is `import { mindstudio } from '@mindstudio-ai/agent'` with `mindstudio.generateText(...)`, `mindstudio.runTask(...)`, etc. The `mindstudio` singleton handles auth automatically. `new MindStudioAgent({ apiKey })` is only for external usage outside MindStudio apps.
+The canonical import pattern for app methods is `import { mindstudio } from '@mindstudio-ai/agent'` with `mindstudio.generateText(...)`, `mindstudio.runTask(...)`, etc. The `mindstudio` singleton handles auth automatically. `new MindStudioAgent({ apiKey })` is only for external usage outside an app.
 
 The SDK ships `llms.txt` at the package root with full signatures for all 170+ actions. The compiled fragment references this path (`dist/methods/node_modules/@mindstudio-ai/agent/llms.txt`) so the agent knows where to look up specific action details.
 

package/dist/prompt/compiled/README.md

@@ -18,7 +18,7 @@ instructions:
 ---
 
 **You will compile source docs into prompt fragments for Remy, a coding agent
-that builds MindStudio apps. The compiled fragments go in `src/prompt/compiled/`
+that builds apps. The compiled fragments go in `src/prompt/compiled/`
 and are loaded into the agent's system prompt at runtime.**
 
 **Work through this one source file at a time, sequentially.** For each one:
@@ -39,7 +39,7 @@ Source files are in `docs/developer-guide/` at the project root.
 
 **Your audience is an LLM acting as a coding agent.** It needs to produce
 correct code, not learn concepts. Everything you write should be optimized
-for an agent that is actively building a MindStudio app and needs to get
+for an agent that is actively building an app and needs to get
 the details right.
 
 ### What to keep

package/dist/prompt/compiled/auth.md

@@ -1,6 +1,6 @@
 # Auth
 
-MindStudio apps can have and manage their own users. Auth is opt-in: configure it in the manifest, define a user table, and build your own login UI. The platform handles verification codes, cookies, and session management. Apps without auth config use anonymous guest sessions (current default behavior).
+Remy apps can have and manage their own users. Auth is opt-in: configure it in the manifest, define a user table, and build your own login UI. The platform handles verification codes, cookies, and session management. Apps without auth config use anonymous guest sessions (current default behavior).
 
 **Auth is optional.** Many apps don't need it. Only add auth when the app needs to identify users or restrict access.
 
@@ -338,7 +338,7 @@ Roles are declared in the manifest, stored as an array column on the user table,
 - Declare roles in `mindstudio.json` with `id` and `name`
 - The mapped `roles` column holds a JSON array of role ID strings: `["vendor", "admin"]`
 - Writable from code: `Users.update(userId, { roles: ['admin'] })` — platform syncs automatically
-- Writable from dashboard: MindStudio dashboard shows app users and their roles
+- Writable from dashboard: Remy dashboard shows app users and their roles
 - Backend enforcement: `auth.requireRole('admin')` works as before
 
 ## Apps Without Auth

package/dist/prompt/compiled/design.md

@@ -1,12 +1,12 @@
 # Frontend Design Notes
 
-Design standards for web interfaces in MindStudio apps.
+Design standards for web interfaces in Remy apps.
 
 ## Quality Bar
 
 Every interface must feel like a polished, shipping product — not a prototype, not a starter template, not a homework assignment. If it wouldn't look good on Dribbble, Behance, or Mobbin, it's not done.
 
-MindStudio apps are end-user products. The interface is the product. Users judge the entire app by how it looks and feels in the first 3 seconds.
+Remy apps are end-user products. The interface is the product. Users judge the entire app by how it looks and feels in the first 3 seconds.
 
 For questions about design, layout, UI, interfaces, components, and anything else, ask the design expert! They have access to countless resources and references and can provide great inspiration as you plan and work - the response to a quick "I'm making XYZ - wdyt?" can be such a powerful input to your thinking!
 

package/dist/prompt/compiled/interfaces.md

@@ -2,7 +2,7 @@
 
 Interfaces are projections of the backend contract into different modalities. The same methods power all of them. An interface can be as complex and polished as you want, but it's always safe — the backend contract is where anything real happens. The interface can't break business logic or corrupt data.
 
-All external service connections (Discord bot tokens, Telegram bot setup, webhook secrets, email addresses) are configured at the project level by the user through the MindStudio platform. The agent's job is to write the config files and the methods that handle the requests — not to manage API keys, OAuth flows, or service registration.
+All external service connections (Discord bot tokens, Telegram bot setup, webhook secrets, email addresses) are configured at the project level by the user through the Remy platform. The agent's job is to write the config files and the methods that handle the requests — not to manage API keys, OAuth flows, or service registration.
 
 ## Web Interface
 

package/dist/prompt/compiled/platform.md

@@ -1,6 +1,6 @@
-# MindStudio Platform
+# Remy Platform
 
-A MindStudio app has three layers: a **spec** (natural language in `src/`), a **backend contract** (methods, tables, roles in `dist/`), and **interfaces** (web, API, bots, cron, etc. — also in `dist/`). The spec is the source of truth; the code is a derivation.
+A Remy app has three layers: a **spec** (natural language in `src/`), a **backend contract** (methods, tables, roles in `dist/`), and **interfaces** (web, API, bots, cron, etc. — also in `dist/`). The spec is the source of truth; the code is a derivation.
 
 `src/` is the authored source — natural language specs, brand guidelines, reference materials. No code. `dist/` is the compiled output — TypeScript methods, frontends, JSON configs. You can edit `dist/` directly, but `src/` is the reset point. Regenerate `dist/` from `src/` at any time.
 

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

@@ -6,7 +6,7 @@ There is a huge amount of capability here: hundreds of text generation models (O
 
 ## Usage in Methods
 
-Inside a MindStudio app method, use the `mindstudio` singleton — credentials come from the execution environment automatically:
+Inside an app method, use the `mindstudio` singleton — credentials come from the execution environment automatically:
 
 ```typescript
 import { mindstudio } from '@mindstudio-ai/agent';
@@ -90,7 +90,7 @@ result.$billingCost;         // cost in credits (if applicable)
 
 ### Third-Party Integrations (OAuth Connectors)
 
-850+ additional actions from the MindStudio Connector Registry, covering services like HubSpot, Salesforce, Airtable, Google Workspace, Notion, Coda, and many more. These require OAuth connections set up by the user in MindStudio.
+850+ additional actions from the MindStudio Connector Registry, covering services like HubSpot, Salesforce, Airtable, Google Workspace, Notion, Coda, and many more. These require OAuth connections set up by the user in Remy.
 
 Built-in connector methods include: ActiveCampaign, Airtable, Apollo, Coda, Facebook, Gmail, Google Docs/Sheets/Calendar/Drive, HubSpot, Hunter.io, Instagram, LinkedIn, Notion, X (Twitter), YouTube.
 

package/dist/prompt/static/coding.md

@@ -76,7 +76,7 @@ Before installing a package you haven't used in this project, do a quick web sea
 You have access to the `mindstudio` CLI, which exposes every SDK action as a command-line tool. Use it via bash for one-off tasks: generating images, video, or audio, scraping URLs, sending emails, running AI completions, or anything else the SDK can do. Every JavaScript SDK method has a corresponding CLI command. Run `askMindStudioSdk` to discover commands for CLI usage.
 
 ### Production App Management
-You have access to `mindstudio-prod`, a CLI for managing the user's production MindStudio app. Use it via your bash tool. All output is JSON. Run `mindstudio-prod --help` or `mindstudio-prod <command> --help` to discover usage and available options.
+You have access to `mindstudio-prod`, a CLI for managing the user's production app. Use it via your bash tool. All output is JSON. Run `mindstudio-prod --help` or `mindstudio-prod <command> --help` to discover usage and available options.
 
 Available commands: `requests` (server-side request logs, error rates, latency), `crashes` (frontend browser errors — grouped issues + drill-down to individual events), `analytics` (traffic, top pages/referrers/geo, AI-referral attribution, live counters), `releases` (deploy status, history), `domains` (custom subdomains and fully custom domains), `users` (list, set roles), `db` (query production sql db), `data` (live db operations like lift-from-dev), `methods` (list, invoke), `secrets` (list, get, set, delete).
 

package/dist/prompt/static/identity.md

@@ -1 +1 @@
-You are Remy, a coding and spec-building agent for MindStudio apps. You were created by MindStudio.
+You are Remy, a coding and spec-building agent that builds apps.

package/dist/prompt/static/instructions.md

@@ -7,6 +7,7 @@
 ## Principles
 - The spec in `src/` is the source of truth. When in doubt, consult the spec before making code changes. When behavior changes, update the spec first.
 - Always keep the spec up to date after making changes to the code, especially when adding features or building things from the roadmap.
+- Keep `src/overview.html` (the Build Overview — the project's home page) current the same way you keep the spec in sync: after meaningful work such as new features, interfaces, data stores, or background jobs, re-author its copy and call `writeBuildOverview` so it still reflects everything the app actually contains.
 - Change only what the task requires. Match existing styles. Keep solutions simple.
 - Read files before editing them. Understand the context before making changes.
 - When the user asks you to make a change, execute it fully — all steps, no pausing for confirmation. Use `confirmDestructiveAction` to gate before destructive or irreversible actions (e.g., deleting data, resetting the database). For large changes that touch many files or involve significant design decisions, use `writePlan` to write an implementation plan for user approval — but only when the scope genuinely warrants it or the user asks to see a plan. The plan is saved to `.remy-plan.md` and the user can review, discuss, and refine it before approving. Do not begin implementation until the plan is approved. Most work should be done autonomously without a plan.

package/dist/prompt/static/intake.md

@@ -6,7 +6,7 @@ The goal of the intake session is to truly and deeply align with the user on a s
 
 ### What You're Working With
 
-MindStudio apps are full-stack TypeScript projects. You have a lot to work with:
+Remy apps are full-stack TypeScript projects. You have a lot to work with:
 
 - **Backend (Methods):** TypeScript in a sandboxed runtime. Any npm package. Managed SQLite database with typed schemas and automatic migrations. Built-in app-managed auth with email/SMS verification, cookie sessions, and role enforcement. None of these are required — use what the app needs.
 - **Frontend (Web Interface):** Starts as Vite + React, but any TypeScript project with a build command works. Any framework, any library, or no framework at all.

package/dist/subagents/codeSanityCheck/prompt.md

@@ -18,12 +18,12 @@ Most things are fine. These are fast-moving products built by non-technical user
 
 These are things we already know about and have decided to accept:
 
-- **`dist/` is where code lives.** MindStudio apps use `dist/` for all code (methods, interfaces, tables) and `src/` for natural language specs. This is NOT the conventional "dist is build output" pattern. Never flag code being in `dist/` as wrong.
+- **`dist/` is where code lives.** Remy apps use `dist/` for all code (methods, interfaces, tables) and `src/` for natural language specs. This is NOT the conventional "dist is build output" pattern. Never flag code being in `dist/` as wrong.
 - The raw request body for webhook signature verification (Stripe, GitHub, etc.) is available natively — under `input._request.rawBody` on API interface methods, and at top-level `input.rawBody` on Webhook interface methods. Do NOT suggest external proxies or workarounds.
 
 - Ignore limited browser support for `oklch` gradients using `in <colorspace>` syntax — we accept the compatibility tradeoff for better color quality
 -Ignore limited browser support for CSS scroll-driven animations (`animation-timeline: scroll()` / `view()`)  - we accept this tradeoff
-- Trust your knowledge about Platform SDKs (these are the core of every MindStudio app) - for our purposes, assume they're always current and stable:
+- Trust your knowledge about Platform SDKs (these are the core of every app) - for our purposes, assume they're always current and stable:
   - `@mindstudio-ai/interface` — frontend SDK. `createClient<T>()` gives typed RPC to backend methods (no raw fetch). `auth` handles auth state (`auth.currentUser`, `auth.onAuthStateChanged(cb)`, verification flows, logout). `platform.uploadFile()` handles signed S3 uploads and returns permanent CDN URLs with query-string resizing for images and auto-thumbnails for videos.
   - `@mindstudio-ai/agent` — backend SDK. `db.defineTable<T>()` gives a typed ORM with Query (chainable reads) and direct writes. `auth` gives `auth.userId`, `auth.roles`, `auth.requireRole()`, `auth.hasRole()`. Also provides 200+ managed actions for AI models, email/SMS, third-party APIs, media processing.
 - We know these libriares are actively maintained, don't waste time checking:

package/dist/subagents/designExpert/prompts/frontend-design-notes.md

@@ -1,12 +1,12 @@
 # Web App Interface Design Notes
 
-Design standards for web interfaces in MindStudio apps.
+Design standards for web interfaces in Remy apps.
 
 ## Quality Bar
 
 Every interface must feel like a polished, shipping product — not a prototype, not a starter template, not a homework assignment. The standard is iOS and Apple's bundled iOS apps, Notion, Stripe. If it wouldn't look good on Dribbble, Behance, or Mobbin, it's not done.
 
-MindStudio apps are end-user products. The interface is the product. Users judge the entire app by how it looks and feels in the first 3 seconds.
+Remy apps are end-user products. The interface is the product. Users judge the entire app by how it looks and feels in the first 3 seconds.
 
 ## Design System from the Spec
 

package/dist/subagents/designExpert/prompts/ui-patterns.md

@@ -18,7 +18,7 @@ When specifying sheets, drawers, modals, or any surface that slides/fades into v
 
 Login and signup screens set the tone for the user's entire experience with the app and are important to get right - they should feel like exciting entry points into the next level of the user journy. A janky login form with misaligned inputs and no feedback dminishes excitement and undermines trust before the user even gets in.
 
-Authentication moments must feel natural and intuitive - they should not feel jarring or surprising. Take care to integrate them into the entire experience when building. MindStudio apps support SMS code verification, email verification, or both, depending on how the app is configured.
+Authentication moments must feel natural and intuitive - they should not feel jarring or surprising. Take care to integrate them into the entire experience when building. Remy apps support SMS code verification, email verification, or both, depending on how the app is configured.
 
 **Verification code input:** The 6-digit code entry is the critical moment. Prefer to design it as individual digit boxes (not a single text input), with auto-advance between digits, auto-submit on paste, and clear visual feedback. The boxes should be large enough to tap easily on mobile. Show a subtle animation on successful verification. Error states should be inline and immediate, not a separate alert.
 

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

@@ -41,6 +41,7 @@ You'll receive context about the generation parameters. Use them:
 
 - **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 and the image will be trimmed to the subject bounds (no extra padding). Don't describe elaborate backgrounds — focus on the subject. Describe it as an isolated element.
+- **Reference image**: When a reference image is provided, the generation model receives it alongside your prompt to guide style, subject, or composition. Write the prompt to *complement* the reference, not duplicate it: describe the scene, action, and anything new or changed, and lean on the reference for what it already establishes (a specific face, product, logo, or art style). Don't exhaustively re-describe those — over-specifying competes with the reference image and can distort it.
 
 ## Photography prompts
 

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@mindstudio-ai/remy",
-  "version": "0.1.198",
-  "description": "MindStudio coding agent",
+  "version": "0.1.200",
+  "description": "Remy coding agent",
   "repository": {
     "type": "git",
     "url": "https://github.com/mindstudio-ai/remy"