79 Commits

This Branch

This Branch

All Branches

• fix: resolve t variable scope issue in class components
  Gemini AI · 2025-12-29 13:06:37 +04:00
  66b31b567b
• feat: consolidate model and provider selector in ai assist header
  Gemini AI · 2025-12-29 12:08:48 +04:00
  3c0dbf574a
• feat: add model selector to ai assist mode
  Gemini AI · 2025-12-29 11:49:07 +04:00
  72814c9496
• feat: complete translations for russian and hebrew across all components
  Gemini AI · 2025-12-29 11:44:55 +04:00
  0589742879
• fix: resolve blank preview in new tabs by using computed state

  - Implemented computed 'currentPreviewData' and 'currentShowCanvas'
  - Uses local state during streaming (isProcessing=true) for immediate feedback
  - Uses tab state (activeTab) when idle for persistence and isolation
  - Updated LiveCanvas and UI to use these computed values
  - Ensures new tabs usually show their own streaming content immediately

  Gemini AI · 2025-12-29 03:53:19 +04:00
  a7f1ea1dc0
• fix: tab preview isolation + remove change log from canvas

  Tab Preview Isolation:
  - LiveCanvas now uses activeTab.previewData directly
  - All canvas UI elements use activeTab state
  - Each tab shows its own preview, not shared state
  
  Change Log Removal:
  - Strip change log text from preview data before rendering
  - Pattern matches [/type:lang] Change Log: and **Change Log:**
  - Change log stays in chat, never rendered in canvas
  - Guards against false positives (app content > 500 chars)

  Gemini AI · 2025-12-29 03:42:15 +04:00
  bbbfb4f90d
• fix: tab isolation + change log in chat not preview

  Tab Isolation:
  - Canvas visibility now uses activeTab.showCanvas directly (not local state)
  - Chat panel width also uses tab state for proper isolation
  - New tabs explicitly have showCanvas: false, previewData: null
  
  Change Log Fix:
  - Updated all AI prompts to emphasize change log OUTSIDE [/PREVIEW]
  - Added example format showing change log after closing tag
  - Change log now appears in chat stream, not rendered in canvas

  Gemini AI · 2025-12-29 03:35:11 +04:00
  fc5fbb1034
• fix: strict tab canvas isolation

  - New tabs explicitly have showCanvas: false and previewData: null
  - Tab sync uses strict boolean check (showCanvas === true)
  - Default tab fallback includes canvas defaults
  - Switching to new tab now properly hides canvas from previous tab
  - Each tab's preview is completely independent

  Gemini AI · 2025-12-29 03:28:20 +04:00
  cfc1ded15e
• feat: independent canvas per tab

  - Added showCanvas to AIAssistTab interface in store
  - Each tab now has its own canvas visibility state
  - Canvas state syncs when switching tabs
  - Closing canvas saves state to tab
  - Clearing history resets canvas state
  - Streaming updates save showCanvas: true to the request's tab
  - Complete tab isolation: history, agent, preview, AND canvas

  Gemini AI · 2025-12-29 03:21:54 +04:00
  4630e23330
• fix: proper tab isolation with updateTabById

  - Added updateTabById function to store for updating specific tabs by ID
  - Capture tab ID at request start (requestTabId)
  - All streaming updates now go to the captured tab ID, not activeTabId
  - Switching tabs during streaming no longer pollutes other tabs
  - Local UI state only updates if still on the same tab
  - Complete isolation: each tab's history/preview/agent is independent

  Gemini AI · 2025-12-29 03:14:07 +04:00
  9996622b12
• fix: complete tab isolation in AI Assist

  - Fixed bug where new tabs were showing content from previous tabs
  - Reset ALL local state when switching tabs:
    - currentAgent, previewData, showCanvas, viewMode
    - assistStep, aiPlan, input, isProcessing
  - Each tab now operates in complete isolation
  - Tab state properly syncs from store on switch

  Gemini AI · 2025-12-29 03:08:06 +04:00
  0e1bdb0c06
• feat: intelligent CSS framework selection for web/app agents

  - Added CSS framework choice to web and app agents across all providers
  - AI can now intelligently select between:
    - Tailwind CSS (default): utility-first, modern designs
    - Windi CSS: faster builds, advanced features
    - Bootstrap: classic, component-based designs
  - Each framework uses its respective CDN for live preview rendering
  - Selection based on design complexity and user requirements

  Gemini AI · 2025-12-29 03:04:09 +04:00
  3a2a07e4a6
• fix: agent selection persistence + premium SEO report design

  - Fixed agent selection to persist when switching tabs (saves to tab state)
  - Added CanvasErrorBoundary to gracefully catch render crashes
  - Added defensive null checks for tab state access
  - Overhauled SEO agent prompt with detailed Google-style dashboard requirements:
    - Dark theme (bg-slate-900)
    - Animated SVG progress rings with stroke-dasharray
    - Color-coded scoring (green/amber/red)
    - Tailwind CDN + Google Fonts (Inter/Roboto)
    - Card layouts with backdrop-blur and subtle borders
    - Key Recommendations section with icons
    - Responsive max-w-4xl layout

  Gemini AI · 2025-12-29 03:00:42 +04:00
  ffa0d2bd91
• feat: multi-tab AI Assist sessions + canvas UX enhancements

  - Implemented multi-tab support for AI Assist (like Gemini/ChatGPT)
  - Added Tab Bar UI with New Chat (+) and Close Tab (x) buttons
  - Synced chat history, agent, and preview data per tab
  - Added BuildingArtifact loading animation during code generation
  - Prevented broken code display during streaming with high-fidelity checklist
  - Updated isPreviewRenderable to detect and skip backend code (Node.js/Express)
  - Enhanced all AI provider prompts for React support and backend simulation
  - Updated SEO Agent to generate ADHD-friendly, modern HTML5 dashboard reports
  - Fixed LiveCanvas try/catch error handling for graceful error displays

  Gemini AI · 2025-12-29 02:49:13 +04:00
  6f3c2949bb
• feat: implement live React/JSX rendering in AI Assist Canvas

  - Enhanced LiveCanvas with Babel Standalone and React/ReactDOM CDNs
  - Implemented automatic import resolution and root mounting for React components
  - Updated parseStreamingContent to identify React code as renderable 'app' artifacts
  - Updated UI to automatically switch to Preview mode for React artifacts
  - Updated all AI provider prompts to officially support and encourage React rendering

  Gemini AI · 2025-12-29 02:35:35 +04:00
  e8904c6b6c
• fix: resolve bracket hallucination and improve tag masking in AI Assist

  - Improved parseStreamingContent to be cực kỳ robust against malformed/bracket-heavy tags
  - Added history transformation to send context in standard Markdown blocks instead of internal tags
  - This prevents the AI from becoming 'tag-confused' and improves surgical edit reliability
  - Refined partial tag hiding for smoother streaming UX

  Gemini AI · 2025-12-29 02:31:28 +04:00
  654dce29e8
• feat: implement surgical iterative edits for AI Assist

  - Updated system prompts for all services to enforce 'ITERATIVE MODIFICATIONS' behavior
  - Instructed AI to maintain design structure and only apply requested changes
  - Added 'Change Log' requirement for text responses
  - Modified AIAssist.tsx to preserve raw code in history for AI context
  - Updated UI rendering to hide code tags while keeping them in context
  - Added 'Applying surgical edits...' status feedback

  Gemini AI · 2025-12-29 02:22:29 +04:00
  695bd74e73
• fix: preserve manually selected agent in AI Assist during streaming
  Gemini AI · 2025-12-29 02:15:31 +04:00
  738ba6cc3a
• fix: hide partial tags in AI Assist and add status feedback UI

  - Refined regex to hide partial [AGENT] and [PREVIEW] tags during streaming
  - Added 'Neural Link Thinking' indicator for initial response phase
  - Added dynamic status badges (e.g., 'Generating web artifact...') during stream
  - Improved robustness of streaming content parser

  Gemini AI · 2025-12-29 02:08:04 +04:00
  0113ba7624
• feat: enhance AI Assist system prompt to prevent refusals and define agent roles
  Gemini AI · 2025-12-29 02:04:41 +04:00
  242f718885
• feat: add 'Same as source' option to slides count

  - Added 'Same as source' to translations in EN, RU, HE
  - Updated 'Number of Slides' dropdown to include 'Same as source' (value 0)
  - Updated Qwen, Ollama, and Z.AI services to handle 0 slide count by instructing AI to match source structure

  Gemini AI · 2025-12-29 01:50:27 +04:00
  ba5eef743e
• fix: revert Qwen model name to 'coder-model' as per user correction
  Gemini AI · 2025-12-29 01:36:24 +04:00
  13ddb7c7c4
• fix: update Qwen model names to correctly supported versions

  - Replaced qwen-coder-plus (which failed with 400) with qwen2.5-coder-32b-instruct
  - Updated available models list with qwen2.5-coder series
  - Corrected store defaults

  Gemini AI · 2025-12-29 01:25:40 +04:00
  c19a31ad61
• fix: call local proxy for Qwen streaming to avoid CORS
  Gemini AI · 2025-12-29 01:08:04 +04:00
  ed3cadf84e
• fix: call Qwen API directly for streaming, not through Next.js route

  - Fixed generateAIAssistStream to call Qwen API directly
  - Uses getEffectiveEndpoint() and getRequestHeaders() for proper auth
  - Added better error logging for debugging
  - Fixed listModels to use getAvailableModels()
  - This fixes the 'Stream request failed' error

  Gemini AI · 2025-12-29 01:00:18 +04:00
  b21ef566b7
• fix: improve AI Assist input visibility and use valid Qwen models

  - Added text-slate-100 dark:text-slate-100 to input for visible text on dark bg
  - Updated Qwen models to use valid names: qwen-coder-plus, qwen-turbo, etc
  - Set qwen-coder-plus as default model for best coding assistance

  Gemini AI · 2025-12-29 00:50:17 +04:00
  f9cbfb0184
• fix: restore AI Assist to use modelAdapter directly

  - Removed intermediate /api/ai-assist call that caused 404 with basePath
  - AI Assist now calls modelAdapter.generateAIAssistStream directly
  - This fixes the HTTP 404 error on the live site

  Gemini AI · 2025-12-29 00:32:35 +04:00
  ac2ba44fbc
• fix(ai-assist): prevent HTML-as-JSON crash; add safeJsonFetch + JSON-only API errors

  - Created lib/safeJsonFetch.ts helper to safely parse JSON responses
  - Updated AIAssist.tsx to use safeJsonFetch with proper Content-Type header
  - Improved app/api/ai-assist/route.ts with requestId tracking and safe body parsing
  - All API responses now return JSON with proper error messages

  Gemini AI · 2025-12-29 00:09:31 +04:00
  42570a14b7
• feat: PromptArch v2.1 - Agentic AI Assist 3-Step Flow + Slides Enhancements

  - Implemented Plan -> Approve -> Generate workflow in AIAssist
  - Added Plan Review Card with architecture/tech stack preview
  - Added Reveal.js Markdown export to SlidesGenerator
  - Created /api/ai-assist and /api/slides orchestration routes
  - Fixed all syntax errors and build issues
  - Maintained premium blue/teal glassmorphism design

  Gemini AI · 2025-12-28 23:48:03 +04:00
  ba1fee3fc7
• fix: additional AI Assist improvements
  Gemini AI · 2025-12-28 21:07:58 +04:00
  37612bbac2
• fix: enhanced HTML detection, fallback mode, blue theme and improved streaming reliability
  Gemini AI · 2025-12-28 12:48:04 +04:00
  8f93e58785
• fix: always default to preview mode for artifacts
  Gemini AI · 2025-12-28 12:12:01 +04:00
  6c84783cc9
• fix: LiveCanvas now always renders when user toggles to Preview mode
  Gemini AI · 2025-12-28 12:07:52 +04:00
  13c45916b7
• fix: enhanced HTML decoding and code fence stripping in AIAssist preview
  Gemini AI · 2025-12-28 11:54:28 +04:00
  d9c6baaa65
• style: emerald theme redesign and ai-assist streaming enhancements
  Gemini AI · 2025-12-28 11:43:13 +04:00
  a2d381cfb7
• feat: complete overhaul of AI Assist with premium WOW level UI and stable preview engine
  Gemini AI · 2025-12-28 11:22:19 +04:00
  d6c2986df8
• feat: AI Assist improvements - Enhanced web/app preview with live iframe - Added structured SEO preview with metrics - Implemented undo and revision (edit user message) capabilities - Fixed undoLast function scope issue
  Gemini AI · 2025-12-28 03:48:12 +04:00
  820e740b9d
• fix: AI Assist - Fix message format and add model selector - Fixed message serialization to string format (fixes 400 API error) - Added model selection dropdown to header - Cleaned message history to plain text before API call - Uses native HTML select for compatibility
  Gemini AI · 2025-12-28 03:43:39 +04:00
  cf26f7a768
• feat: Add AI Assist module - Conversational intelligence with specialized agent switching - Real-time preview for code, design, content, and SEO - Multi-agent routing logic (Content, SEO, SMM, PM, Code, Design) - High-end UI with chat thread and live preview sandbox - Integrated with Ollama, Qwen, and Z.AI services
  Gemini AI · 2025-12-28 03:35:29 +04:00
  bc627479f1
• feat: Include direct product URLs in Market Research - Updated MarketResearchResult type to include optional URLs - Enhanced AI prompts to extract competitor product links - Added 'View Product' links to the Price Matrix UI - Responsive links with protocol verification
  Gemini AI · 2025-12-28 03:23:01 +04:00
  8f0a65eee1
• feat: Add progress bar and AI thoughts to Market Research - Added dynamic progress bar below research parameters - Implemented cycling AI thought messages during processing - Localized thought messages in English, Russian, and Hebrew - Enhanced visual feedback for deep analysis process
  Gemini AI · 2025-12-28 03:11:06 +04:00
  7625b20aa3
• feat: Add AI Market Research module - Introduced Automated Market Analysis and Competitive Intelligence feature - Added 'Market Research' view to Sidebar and main navigation - Implemented MarketResearcher component with price comparison matrix and feature tables - Updated Ollama, Qwen, and Z.AI services to support market research generation - Added localized translations in English, Russian, and Hebrew
  Gemini AI · 2025-12-28 03:02:00 +04:00
  792685ca6b
• feat(google-ads): Add 'Special Instructions' support - Users can now provide custom preferences (tone, focus, etc.) for AI generation - Added 'Special Instructions' field to Google Ads Generator UI (below Target Audience) - Localized labels and placeholders in English, Russian, and Hebrew - Integrated instructions into AI prompts across Ollama, Qwen, and Z.AI providers
  Gemini AI · 2025-12-28 02:51:53 +04:00
  2c7b233bca
• feat(magic-wand): Create 'noob-friendly' Google Ads setup guide - Magic Wand now generates specific Headlines, Descriptions, and Keywords for each strategy - Added visual 'Google Ads Setup Guide' cards within each strategic direction - Implemented click-to-copy functionality for headlines and descriptions - Added beginner tips explaining exactly where to paste content in Ads Manager - Updated types and AI prompts across all providers (Ollama, Qwen, Z.AI)
  Gemini AI · 2025-12-28 02:44:03 +04:00
  9e74fcba67
• style(google-ads): Fix results card opacity and revamp UI with premium styling - Fixed bug where results card stayed dimmed at opacity-50 when magic wand results were shown - Added glassmorphism, depth, and vibrant gradients to market analysis cards - Implemented realistic Google Search preview for ad copies - Enhanced competitor insights with better hierarchy and spy-themed data blocks - Modernized campaign structure with a high-contrast dark theme
  Gemini AI · 2025-12-28 02:34:05 +04:00
  f109e79c0b
• feat(google-ads): Add product URLs and animated progress messages - Products now support custom URLs for linking ads and AI research - Added animated progress messages with fun AI statements during generation - Progress bar shows cycling status updates in EN/RU/HE - Enhanced UX with bouncing dots and gradient progress indicator
  Gemini AI · 2025-12-28 02:23:30 +04:00
  932de22c97
• fix(i18n): Fix syntax errors and add missing translation keys - Restored broken useStore destructuring in PromptEnhancer, UXDesignerPrompt, ActionPlanGenerator, SlidesGenerator - Fixed GoogleAdsGenerator scope issue with renderMagicWandSectionContent - Added missing inputLabel keys to promptEnhancer for all languages - Fixed t.resultTitle to t.enhancedTitle in PromptEnhancer
  Gemini AI · 2025-12-28 02:00:42 +04:00
  238a576cb8
• feat(i18n): Complete localization for Settings, Action Plan, Slides, and Google Ads generators - Added actionPlan and slidesGen translation sections for EN/RU/HE - Localized ActionPlanGenerator with dynamic titles, labels, and quick notes - Localized SlidesGenerator with theme, audience, and animation settings - Added RTL support and text-start alignment for Hebrew - Standardized AI Provider and Model labels across all generators
  Gemini AI · 2025-12-28 01:46:32 +04:00
  b859d77307
• refactor: Google Ads Gen UI to match PRDGenerator design pattern

  - Replaced flashy styling with simple two-column card layout
  - Added collapsible sections for Keywords, Ad Copy, Campaigns, Implementation
  - Consistent typography and spacing with other generators
  - Removed all gradients, glassmorphism, and excess animations

  Gemini AI · 2025-12-28 01:24:17 +04:00
  8733b885d4
• fix: Google Ads parsing and UI/UX refinement

  - Added robust JSON extraction logic to handle AI responses with markdown formatting.
  - Overhauled Google Ads Gen UI with glassmorphism and premium aesthetics.
  - Harmonized main content styling with the sidebar design ecosystem.

  Gemini AI · 2025-12-28 01:19:41 +04:00
  841eaab5a7