green-app

As a user (trader or admin), I want to log in using my credentials on a styled login page that matches the Login (v1) mock design so that I can securely access my personalized dashboard. The page should include email/password fields, validation feedback, and a link to sign up. Page is reached from: Landing page. Navigates to: Admin Dashboard (for admin) or Broker Connect (for new traders) or Dashboard (for returning traders).

As a trader, I want to connect my Quotex broker account by entering my broker ID on the Broker Connect page so that the platform can retrieve my trading data and provide AI-powered predictions. The page should include a broker ID input, connection status indicator, and validation messages matching the Broker Connect (v2) mock design. Page is reached from: Login (first-time traders). Navigates to: Dashboard upon successful connection.

As an admin, I want to view a comprehensive admin dashboard with platform KPIs, active user counts, system health indicators, recent activity feed, and quick links to user management and AI model monitoring so that I can oversee the platform. Implement based on Admin Dashboard (v1) mock design with data tables and stat cards. Page is reached from: Login (admin role). Navigates to: AI Models, Users, Settings.

As a developer, I want FastAPI admin endpoints to list all users, search and filter by status or broker connection, view user details, suspend or reinstate accounts, and flag compliance issues so that admins can effectively manage the platform user base. Includes /admin/users, /admin/users/{id}, /admin/users/{id}/suspend, /admin/users/{id}/reinstate endpoints.

As a developer, I want FastAPI endpoints to validate and store a trader's Quotex broker ID, verify broker connectivity, retrieve broker account metadata, and manage broker connection status so that traders can link their Quotex accounts to the platform. Includes /broker/connect, /broker/verify, /broker/status, /broker/disconnect endpoints.

As a prospective trader, I want to see a compelling landing page with hero section, feature highlights, call-to-action buttons (Get Started, Login), and branding so that I understand the value of the green-app trading prediction platform before signing up. Implement based on the existing Landing (v1) JSX mock design. Page links to: Login page and Signup page.

As a developer, I want FastAPI endpoints to manage notification preferences, send push/email notifications for trading signals and alerts, and retrieve notification history so that traders receive timely updates. Includes /notifications/preferences, /notifications/history, /notifications/send, /notifications/mark-read endpoints.

As an AI engineer, I want to implement a Langchain-based signal generation pipeline using LiteLLM to route requests to GPT 5.2 and Claude 4.5 Opus that analyzes live market data from connected Quotex broker accounts and produces structured BUY/SELL/HOLD signals with confidence scores so that traders receive AI-powered trade recommendations. Includes prompt engineering, chain construction, model routing, and output parsing.

As an admin, I want to view, search, filter, and manage all registered traders on the Users page including account status, broker connection status, compliance flags, and the ability to suspend or review accounts so that I can maintain platform integrity. Implement based on Users (v1) mock design with data grid and action menus. Page is reached from: Admin Dashboard.

As an AI engineer, I want to integrate observability tooling into the LiteLLM and Langchain pipeline to trace all AI model calls, log input/output tokens, latency, cost, and accuracy outcomes so that the admin AI Models page has real data to display and model performance can be continuously monitored and improved.

As a developer, I want FastAPI endpoints to generate, retrieve, and stream live AI trading signals and predictions for connected traders so that the dashboard can display real-time actionable insights. Includes /signals/live, /signals/history, /predictions/generate, /predictions/latest endpoints with WebSocket support for live streaming.

As an AI engineer, I want to implement a market analysis engine using Langchain and LiteLLM that evaluates multiple trading markets simultaneously, scores them by predicted volatility and opportunity, and automatically selects the top recommended markets for each trader so that the auto-select feature provides intelligent market curation. Includes multi-market scoring prompts, ranking logic, and scheduled refresh via background tasks.

As a trader, I want to view my main dashboard with AI prediction scores, live trading signals, account summary, and quick navigation to key features so that I can make informed trading decisions at a glance. Implement based on Dashboard (v1) mock design including cards, signal indicators, and chart placeholders. Page is reached from: Broker Connect or Login. Navigates to: Markets, History, Settings.

As a developer, I want FastAPI endpoints to retrieve a trader's historical trades, prediction accuracy records, profit/loss summaries, and paginated trade logs so that the History page can display comprehensive performance data. Includes /history/trades, /history/predictions, /history/summary, /history/export endpoints with filtering and pagination.

As a developer, I want FastAPI endpoints to retrieve available trading markets, AI-recommended market selections, market trend data, and toggle auto-select preferences so that traders can view and act on market intelligence. Includes /markets/list, /markets/trends, /markets/auto-select, /markets/recommendations endpoints.

As a developer, I want FastAPI admin endpoints to retrieve AI model performance metrics, accuracy scores, active model status, update model configuration parameters, and toggle active models (GPT 5.2, Claude 4.5 Opus) so that admins can monitor and control AI behavior. Includes /admin/ai-models, /admin/ai-models/{id}/metrics, /admin/ai-models/{id}/config, /admin/ai-models/{id}/toggle endpoints.

As an admin, I want to monitor and update AI model configurations (GPT 5.2, Claude 4.5 Opus) including model performance metrics, accuracy scores, active model toggles, and configuration parameters on the AI Models page so that I can ensure optimal prediction quality. Implement based on AI Models (v1) mock design with model cards and metric charts. Page is reached from: Admin Dashboard.

As a trader, I want to view available trading markets, AI-recommended auto-selected markets, and market trend indicators on the Markets page so that I can identify the best opportunities. Implement based on Markets (v1) mock design with market cards, trend charts, and an auto-select toggle. Page is reached from: Dashboard. Navigates back to: Dashboard.

As a trader, I want to manage my account settings including notification preferences, broker connection details, display preferences, and account security on the Settings page so that I can customize my trading experience. Implement based on Settings (v1) mock design with tabbed sections and toggle controls. Page is reached from: Dashboard.

As a trader, I want to view my trading history including past predictions, outcomes, profit/loss summaries, and signal accuracy statistics on the History page so that I can evaluate performance over time. Implement based on History (v1) mock design with filterable table, summary cards, and export options. Page is reached from: Dashboard.