mammoth-ai

As a backend developer, I want to build the user profile and settings API so that users can retrieve and update their preferences and account data. Implement GET /api/profile/:userId, PATCH /api/profile/:userId for bio and avatar updates, GET /api/settings/:userId, and PATCH /api/settings/:userId for notification preferences, privacy toggles, and companion personality sliders.

As a UI/UX designer, I want to establish a unified color palette, typography scale, and theme tokens (primary, accent, background, surface, text colors) so that every page feels cohesive and on-brand for a flirty, playful AI girlfriend app. Define CSS variables or design tokens, apply them globally, and validate against the Home page design baseline from t1.

As a backend developer, I want to build the admin user management API so that administrators can view, search, filter, suspend, and delete user accounts. Implement GET /api/admin/users with query params for search, filter, and pagination; GET /api/admin/users/:id for full user detail; PATCH /api/admin/users/:id/status to suspend or reactivate; and DELETE /api/admin/users/:id with cascade handling.

As a backend developer, I want to build the onboarding and companion configuration API so that user customization choices are persisted and retrievable. Implement POST /api/companions to create a companion profile with name, appearance attributes, and personality traits; GET /api/companions/:userId to retrieve config; and PATCH /api/companions/:id to update appearance or personality settings.

As a frontend developer, I want to implement the Landing page from its JSX v1 design so that new visitors get an engaging first impression of the AI girlfriend app. Implement hero section, value proposition copy, call-to-action buttons (Sign Up, Login), and any animated or visual elements. Wire navigation links to Login and Signup routes.

As a backend developer, I want to build the chat messaging API so that user messages are stored and routed to the AI engine for responses. Implement POST /api/chat/message to accept user input, persist to the messages table, trigger the AI engine, and stream or return the AI reply; GET /api/chat/history/:userId for paginated message history; and POST /api/chat/react to store emoji reactions on messages.

As a backend developer, I want to build the admin analytics and reporting API so that administrators can retrieve aggregated platform metrics and export reports. Implement GET /api/admin/reports/overview for KPI summaries, GET /api/admin/reports/users for user growth time-series data, GET /api/admin/reports/messages for message volume stats, and GET /api/admin/reports/export to generate and return a CSV or PDF report file.

As a frontend developer, I want to implement the Admin Dashboard page from its JSX v1 design so that administrators have a high-level overview of platform health and activity. Render KPI cards (total users, active sessions, messages today, AI response rate), recent activity feed, quick-nav links to Users, AI Config, Reports, and Settings sections.

As a frontend developer, I want to implement the Signup page from its JSX v1 design so that new users like Mikey can create an account quickly. Implement name, email, password, and confirm-password fields, inline validation, terms acceptance checkbox, submit handler wired to the registration API, and redirect to Onboarding upon success.

As a frontend developer, I want to implement the Login page from its JSX v1 design so that returning users can securely sign in to access their companion. Implement email/password form fields, form validation feedback, submit handler wired to the auth API, error state display, and a link to the Signup page.

As a frontend developer, I want to implement the Admin Users page from its JSX v1 design so that admins can monitor, search, filter, and manage all registered user accounts. Implement a searchable/sortable data table with user rows (avatar, name, email, status, join date, message count), row-level action menus (view profile, suspend, delete), and pagination.

As an AI engineer, I want to integrate a conversational LLM (e.g., OpenAI GPT-4o or similar) into the chat pipeline so that user messages receive intelligent, context-aware replies from the AI companion. Implement the AI service layer that accepts a message payload and conversation history, calls the LLM API with a constructed prompt, handles streaming responses, manages token limits, and returns the AI reply to the chat API.

As a frontend developer, I want to implement the first Onboarding step (Customize AI) from its JSX v1 design so that users can begin personalizing their AI companion. Render the step indicator, AI name input, relationship style selector, and a preview of the companion card. Wire form state to the onboarding context and advance to the next step on submission.

As a frontend developer, I want to implement the Admin Reports page from its JSX v1 design so that admins can view analytics and export data for business insights. Render date-range picker, line/bar charts for user growth and message volume, top active users table, AI response quality metrics, and an export to CSV/PDF button.

As a frontend developer, I want to implement the Admin Settings page from its JSX v1 design so that admins can configure platform-wide settings such as feature flags, subscription tiers, email templates, and API key management. Implement grouped settings sections with form controls, save confirmations, and danger zone actions.

As a frontend developer, I want to implement the Admin AI Config page from its JSX v1 design so that admins can tune the global AI personality, tone, and response behavior. Implement model selector, temperature/creativity sliders, system prompt editor, tone preset toggles (flirty, playful, sweet, bold), a live test chat panel, and save/publish controls.

As an AI engineer, I want to build a dynamic system prompt builder that translates the user's companion personality and tone settings into LLM instructions so that each user's AI girlfriend feels uniquely tailored to their preferences. Implement a PromptBuilder class that composes a base system prompt from companion name, relationship style, personality trait weights, and tone presets, and injects it at the start of every LLM call.

As a frontend developer, I want to implement the Onboarding Set Appearance step from its JSX v1 design so that users can visually customize their AI girlfriend's look. Implement avatar/style selectors (hair, eyes, outfit theme), a live companion preview panel, and navigation between onboarding steps. Persist selections to the onboarding state.

As an AI engineer, I want to fine-tune and prompt-engineer the AI companion to consistently produce flirty, sexy, and playful responses appropriate to the app's persona while staying within content safety boundaries. Implement a curated set of few-shot examples for flirty dialogue, a response post-processor that injects playful emojis or pet names, and a content moderation layer that filters disallowed content before delivery to the user.

As a frontend developer, I want to implement the Onboarding Set Personality step from its JSX v1 design so that users can dial in their companion's personality traits (flirty, playful, caring, etc.). Implement trait sliders or tag selectors, tone preview quotes that update dynamically, a finish/confirm button that submits all onboarding data to the API, and redirects to the Home page.

As a frontend developer, I want to implement the Home page from its JSX v1 design so that users land on an immersive companion experience after onboarding or login. Render the companion avatar/card with her name and mood indicator, a greeting message, quick-action buttons (Start Chat, View Profile, Settings), and bottom navigation. Fetch companion profile from API on mount.

As a frontend developer, I want to implement the Profile page from its JSX v1 design so that users can view and manage their account details and companion configuration. Display user avatar, username, account stats (messages sent, relationship level), companion summary card, and edit profile CTA. Wire data fetching to the profile API endpoint.

As a frontend developer, I want to implement the Chat page from its JSX v1 design so that users can have real-time flirty and playful conversations with their AI companion. Implement the message thread list with user and AI message bubbles, typing indicator animation, message input with send button, emoji/reaction picker, and scroll-to-bottom behavior. Wire to the chat API and handle streaming or polling AI responses.

As a frontend developer, I want to implement the user Settings page from its JSX v1 design so that users like Mikey can control their account preferences, notification settings, companion personality tuning, and privacy options. Implement toggle switches, dropdowns, save/cancel actions, and wire each section to the appropriate settings API endpoints.