cedar-chatgpt
chatgpt app
Comments (0)
Sign in to leave a comment
Project Tasks17
Implement Theme Color System
As a user I want a consistent visual experience across the app so that the interface feels cohesive, requiring implementation of CSS custom properties (--primary: #FF6F61, --bg: #F5F9FF, --surface: #FFFFFF, --text: #2C3E50, --text_muted: #BDC3C7, --border: rgba(189,195,199,0.4)), a typography scale using Inter font, reusable component tokens, and light/dark theme switching support. This must be completed before any page UI tasks begin.
Save Conversation Backend API
As a user I want to save a conversation so that I can come back to it later, requiring FastAPI endpoints to create, update, and soft-delete saved conversation records in MySQL (POST /api/conversations, PUT /api/conversations/{id}, DELETE /api/conversations/{id}), associating them with the authenticated user, storing full message history with timestamps, and supporting star/tag metadata for History page filtering.
Integrate Voice-to-Text AI
As a user I want my voice recordings accurately transcribed so that the AI understands what I said, requiring integration of OpenAI Whisper (or cloud STT fallback) into the voice chat backend pipeline, including audio format normalization to WAV/MP3, chunked streaming transcription support, transcription error handling with user-facing feedback, and latency optimization targeting under 2 seconds for typical utterances.
Theme Preference Backend API
As a user I want my theme settings saved to my account so that my preferences persist across devices and sessions, requiring FastAPI endpoints GET /api/user/preferences and PUT /api/user/preferences to read and update per-user theme (light/dark/coral), AI model preference, and notification settings in MySQL, callable from the Settings page frontend.
Integrate LiteLLM AI Routing
As a user I want my messages routed to the best available AI model so that I always get high-quality responses, requiring LiteLLM integration in the FastAPI backend to dynamically route requests to GPT 5.2, Claude 4.5 Opus, or Gemini 3 Pro based on query type and user model preference, with unified streaming response formatting, per-provider error fallback, and request logging for the API supervisor layer.
Conversation History Backend API
As a user I want to retrieve my saved conversations so that I can view and continue past chats, requiring FastAPI endpoints: GET /api/conversations (paginated list with search query, date-range filter, starred filter, tag filter), GET /api/conversations/{id} (full conversation with messages), POST /api/conversations/bulk (bulk star/delete for BulkActionBar), and GET /api/conversations/export (CSV/JSON export for selected IDs).
Chat Conversation Backend API
As a user I want my messages to be processed and responded to by the AI so that I can have a real conversation, requiring FastAPI endpoints to accept user messages, manage conversation session state, stream AI responses via Server-Sent Events back to the client, persist conversation data to MySQL, and integrate with LiteLLM routing. Endpoints: POST /api/chat/message, GET /api/chat/{session_id}/stream.
Scaffold Frontend App Structure
As a user I want fast, reliable navigation between all pages so that I can move seamlessly through the app, requiring React Router v6 setup with routes for Landing, Login, Signup, Home, Chat, History, and Settings pages, a shared layout component, reusable UI primitives (Button, Input, Card, Modal), and folder structure aligned with the v2 design system. Remove any scaffold pages not present in the design (e.g. old welcome/dashboard pages).
Voice Chat Backend API
As a user I want to send voice messages and receive spoken or text responses so that I can chat hands-free, requiring a FastAPI endpoint POST /api/chat/voice to accept audio blob uploads, validate audio format, coordinate with the STT service for transcription, route the transcribed text through the LiteLLM chat pipeline, and return the AI text response (and optionally TTS audio).
Build Landing Page UI
As a user I want to see an engaging landing page so that I understand the platform's value proposition and can navigate to login or signup. Implement the existing v2 JSX design (NavBar, Hero with floating chat bubbles, FloatingBubblesShowcase, FeatureHighlights, HowItWorks, ModelShowcase, VoiceChatCallout, Testimonials, PersonalizationPreview, CallToAction, Footer) as a fully functional React page using the established theme. Animate floating bubble canvas at page root level. Links navigate to /Login and /Signup.
Build Login Page UI
As a user I want to log in to my account so that I can access my saved conversations and personalized settings. Implement the existing v2 JSX Login design as a fully functional React page with email/password fields, inline validation, error state handling, and redirect to /Home on success. Include a link to /Signup for new users. Styled with the app theme including coral accent CTA button and frosted-glass card surface.
Build Signup Page UI
As a new user I want to create an account so that I can start using the AI chat platform. Implement the existing v2 JSX Signup design as a fully functional React page with name, email, and password fields, inline validation, success/error feedback, and redirect to /Home upon successful registration. Include a link to /Login for existing users. Styled with the coral-accent theme and frosted-glass card.
Build API Supervisor Service
As a user I want the platform to remain stable and responsive under varying load so that my experience is never degraded, requiring a backend supervisor layer that manages pre-resolved user entities, handles rate limiting and quota tracking per LLM provider (GPT/Claude/Gemini), enforces request validation, caches frequently resolved entities in Redis or in-memory store, and exposes GET /api/health and GET /api/metrics endpoints for operational visibility.
Build Home Page UI
As a logged-in user I want a central home dashboard so that I can quickly start a new chat, access conversation history, or go to settings. Implement the existing v2 JSX Home design (NavBar with coral CTA, Hero with animated floating bubbles, FeatureHighlights, HowItWorks, ModelShowcase, VoiceChatCallout, Testimonials, PersonalizationPreview, CallToAction, Footer) as a fully functional React page. Navigation links route to /Chat, /History, and /Settings.
Build Chat Page UI
As a user I want to send and receive AI messages in a chat interface so that I can have a natural conversation. Implement the existing v2 JSX Chat design as a fully functional React page including: message thread view with user/AI bubble distinction, streaming response rendering, text input with send action, voice input trigger button, AI model selector badge, save conversation control, and sidebar with conversation history. Styled with coral accent and frosted-glass surfaces.
Build Settings Page UI
As a user I want to customize my app experience so that the interface matches my preferences. Implement the existing v2 JSX Settings design as a fully functional React page including: theme selection controls (light/dark/coral accent), preference persistence to the backend theme API, live preview of theme changes, profile management section, notification preferences, and AI model preference selector. Styled consistently with the app design system.
Build Conversation History Page
As a user I want to browse my saved conversations so that I can revisit past interactions. Implement the existing v2 JSX History design as a fully functional React page including: NavBar, PageHeader with export/clear controls, SearchFilterBar with filter chips (All, Today, This Week, Starred), ConversationSidebar with tag/collection filters, ConversationList with card hover actions (star, share, delete), BulkActionBar for multi-select operations, PaginationBar, and Footer. Connects to the conversation history and save APIs.
No comments yet. Be the first!