How tecnhotron/singularity works
Project type and purpose
• This is a full-stack chat application (“Singularity”) with:
• A static frontend (HTML/CSS/JS)
• A Flask-based REST API backend
• Firebase Firestore as the data store
• Integration with the Gemini AI API for generating chat responses
File structure
• Top-level files
• Dockerfile: containerization of the service
• server.py (or app.py in README): main Flask application
• requirements.txt: Python dependencies
• Static assets
• index.html, login.html, register.html: client entry points
• css/: styling
• js/: client logic modules (login, register, chat script, firebase-config, document-tool, image-tool, utilities)
• External config
• firebase-key.json: service account credentials for Firestore
• firebase-config.js: client Firebase SDK config
• README.md: describes backend setup, endpoints, and features
README for insights
• Lists REST endpoints for chats, messages, settings, gems
• Describes use of Firestore, Gemini API, and Prompt Gems
• Notes custom chat-creation behavior in frontend
Main components
a. Frontend (Browser)
• HTML pages + CSS
• JS modules: authentication, chat UI, document/image generation, API key management
• Firebase Web SDK for client-side config (optional)
b. Backend (Flask App)
• API layer with endpoints under /api/*
• Request handlers for chats, messages, settings, gems, new-chat
• Integration module for calling Gemini API
• Firestore client library for data persistence
c. Database
• Firebase Firestore: collections for users, chats, messages, gems, settings
d. External Service
• Gemini AI API (Google AI Studio) for AI-powered responses
Relationships and interactions
• User interacts with the frontend (login, register, chat UI)
• Frontend sends HTTP requests (AJAX/fetch) to Flask API
• Flask API:
• Reads/writes chat records in Firestore
• Reads/writes user settings and prompt gems
• Calls Gemini API to generate AI responses
• Returns JSON payloads to the client
• Optionally, client may use Firebase SDK for real-time updates (if applicable)
Architectural patterns and design principles
• Client–server separation: UI logic in JavaScript, business logic in Flask
• RESTful API design for modular, stateless endpoints
• Layered architecture: Presentation (frontend), Application (Flask), Data (Firestore), External (Gemini)
• Extensibility points: Prompt Gems endpoint, document/image tool modules
Relevant technologies/frameworks
• Frontend: Vanilla JS, Firebase Web SDK, HTML5, CSS3
• Backend: Python 3.8+, Flask, firebase-admin SDK, requests (for Gemini)
• Deployment: Docker (Dockerfile)
• Database: Google Firestore
• AI Service: Google Gemini API