Project: Automated One-Click Sign-In & Page Data Extraction System Objective Replace a manual sign-in workflow with a single-button automation that logs into a third-party web app, extracts its page schema and API calls, and makes the data accessible directly through our own website. Scope of Work Step 1 — One-Click Login Automation • Create a secure backend API endpoint that listens for a button click. • On click, it should: o Fetch the target URL, login ID, and password from our internal database. o Use a headless browser or automation framework (Selenium, Playwright, or Puppeteer) to: Open the target web application. Auto-fill credentials into the HTML login form. Submit the form and confirm successful login (e.g., via page redirect or success indicator). • Include robust error handling for: o Wrong credentials o Time-outs / CAPTCHA o Changed page structure Step 2 — Page Schema Extraction • Once logged in, extract the complete DOM / HTML schema of the landing page. • Parse and store: o Page elements and structure o Input fields, forms, and buttons o Relevant identifiers and attributes (names, IDs, types, values) • Save this structured schema in our own database for internal replication (JSON preferred). Step 3 — API Call & Data Flow Replication • Intercept and log all third-party API calls made by that page (XHR / fetch / WebSocket, etc.). • Replicate those API endpoints and their payloads through our own backend, so future data posting happens from our site directly using: o The stored login credentials o Cookies / session tokens from the automated sign-in • Provide a secure mechanism to refresh tokens and maintain session persistence. Deliverables • Fully functional prototype (local or small cloud instance) • Well-commented source code • Setup & usage documentation (README) • Step-by-step installation guide for our internal team Preferred Tech Stack • Language: Python (FastAPI / Flask) or Node.js (Express) • Automation: Playwright / Puppeteer / Selenium (any headless mode) • Database: MySQL (credentials + schema storage) • Deployment: Docker or standalone script • Bonus: Support for multi-account credential sets Additional Notes • Security is crucial: credentials and cookies must never be exposed in logs. • The code should be modular for future extensions (e.g., multi-site login or session reuse). • Flexibility to adapt to both static and dynamically rendered pages (JS frameworks).