distretto

distretto — Your AI-Powered Browser Automation Assistant Automate any browser task using natural language. Just describe what you want, and distretto handles the rest. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ WHAT IS DISTRETTO? distretto is an AI browser automation agent that uses GPT-4 Vision to see and interact with web pages on your behalf. Instead of writing scripts or recording macros, simply tell distretto what you want to do in plain English — click buttons, fill forms, scroll pages, or extract information. Think of it as having a helpful assistant that can navigate any website for you. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ KEY FEATURES • Click — Click any button, link, or interactive element. Just describe it: "Click the Sign In button" or "Click the third item in the list." • Type — Fill out forms and input fields automatically. "Type my email in the login field" or "Enter 'Hello World' in the search box." • Scroll — Navigate through long pages. "Scroll down to the pricing section" or "Scroll to the bottom of the page." • Drag — Move sliders, drag-and-drop elements, and interact with complex UI components like canvas drawing tools. • Summarize — Extract key information from articles, product pages, or complex content. "Summarize this article" or "What are the main features listed on this page?" • Solve — Complete quizzes, puzzles, and interactive challenges. distretto understands questions and selects the correct answers. • Tutor — Get explanations and step-by-step guidance while browsing. "Explain what this error message means" or "Help me understand this form." ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ HOW IT WORKS distretto operates in a simple loop: Observe — Captures a screenshot of your current browser tab Think — AI analyzes the page and decides the best action Act — Executes the action with visual feedback Repeat — Continues until your task is complete The entire process happens in a side panel that stays open while you browse. You can watch distretto work in real-time and stop it at any point. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ EXAMPLE USE CASES Productivity • "Fill out this contact form with my name and email" • "Find the download button and click it" • "Navigate to the settings page" Research • "Summarize the key points on this page" • "What pricing plans are available?" • "List all the features mentioned" Learning • "Help me complete this quiz" • "Explain what each option does" • "Guide me through this tutorial" Testing • "Click through each menu item" • "Test all the links on this page" • "Fill the form with test data" ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ TRANSPARENT PRICING distretto itself is completely free. You only pay for the AI API usage: • GPT-4o Mini (Recommended): ~$0.001-0.003 per action • GPT-4o: ~$0.01-0.03 per action A typical 20-step automation session costs approximately $0.02-0.06 with GPT-4o Mini. Smart Cost Optimization: distretto uses OpenAI's prompt caching feature, which reduces costs by approximately 75% on multi-step tasks. The system prompt is cached after the first request, so subsequent actions in the same session are significantly cheaper. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ PRIVACY & SECURITY Your privacy is paramount: • Your API key stays local — It's stored only in Chrome's secure local storage and is never transmitted to any server except OpenAI's API. • No data collection — distretto doesn't collect, store, or transmit your browsing data, screenshots, or conversation history to any third-party servers. • Direct API calls — All AI requests go directly from your browser to OpenAI. There's no middleman server. • Open inspection — As a browser extension, you can inspect exactly what data is being sent in Chrome DevTools. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ GETTING STARTED Install distretto from the Chrome Web Store Click the extension icon to open the side panel Click the Settings (gear) icon Enter your OpenAI API key (get one at platform.openai.com/api-keys) Click "Test Connection" to verify Start automating! ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ REQUIREMENTS • Chrome browser version 116 or later • OpenAI API key with GPT-4 Vision access ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ AVAILABLE COMMANDS distretto understands natural language, but here are the core capabilities: • navigate("url") — Go to any website • click("target") — Click elements by their visible text • type("text", "field") — Type into input fields • scroll(direction, amount) — Scroll the page • scroll_to("target") — Scroll to make an element visible • drag(x1, y1, x2, y2) — Drag from one point to another • hover("target") — Hover over elements • press_key("key") — Press keyboard keys (Enter, Tab, Escape, etc.) • wait(ms) — Wait for page updates ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ TIPS FOR BEST RESULTS • Be specific about what you want to click: "Click the blue Submit button" works better than just "Click submit" • For forms, specify which field: "Type 'john@example.com' in the email field" • Use scroll commands to reveal hidden content before asking distretto to interact with it • For multi-step tasks, you can give one instruction at a time or describe the full workflow • Watch the visual feedback — distretto shows exactly where it's clicking