Turn your web browsing into actionable insights with AI that understands what you're actually trying to do.
A Chrome extension that records your web interactions and uses AI to analyze your tasks, understand your intent, and suggest optimizations. Perfect for UX research, process documentation, and automation planning.
🤖 Dual AI Analysis
- Chrome AI: Private, local analysis using Chrome's built-in Gemini Nano
- Claude AI: Advanced cloud analysis with Anthropic's Claude
- Roasting Mode: Get hilarious roasts about your task inefficiencies 🔥
📊 Smart Task Understanding
- Records every click, type, scroll, and navigation
- Analyzes your intent behind each action
- Suggests workflow optimizations
- Generates automation-ready instructions
🔒 Privacy First
- All data stays on your device (unless using cloud AI)
- Passwords automatically masked
- No tracking, no analytics, no data collection
# Clone this repo
git clone https://github.com/yourusername/user-intent-extension.git
cd user-intent-extension
# Install dependencies
npm install
# Build the extension
npm run build- Open Chrome and go to
chrome://extensions/ - Enable "Developer mode" (top right toggle)
- Click "Load unpacked" and select the
distfolder - Click the extension icon to open the side panel
- Hit "Start Mark" to begin recording a task
- Do your thing on any website
- Hit "Stop Mark" when done
- Get instant AI analysis of what you just did! 🎉
- Requirements: Chrome 131+
- Setup: Nothing! It just works locally and privately
- Get an Anthropic API key
- Create a
.envfile:VITE_ANTHROPIC_API_KEY=your_api_key_here
- Rebuild:
npm run build - Switch to Claude in the Debug tab
- Intent Recognition: "User is trying to book a flight"
- Step Breakdown: Each action with its purpose explained
- Optimization Tips: "You could bookmark this page for faster access"
- Automation Recipe: Natural language instructions for browser automation
Enable in Debug tab for hilarious takes on your productivity:
"😂 Took 12 clicks to do what could've been done with Ctrl+F... efficiency is not your strong suit 🤦♂️"
- Export complete interaction data
- Analysis history with timestamps
- Compressed JSON format for easy sharing
# Development with hot reload
npm run dev
# Lint and format
npm run lint
# Build for production
npm run build- UX Research: Understand how users really interact with your app
- Process Documentation: Auto-generate step-by-step guides
- Automation Planning: Get prompts ready for Playwright/Selenium
- Training Materials: Show the "why" behind each interaction
- Bug Reports: Detailed reproduction steps with context
- Workflow Optimization: Find inefficiencies in complex tasks
- Frontend: TypeScript + Chrome Extensions Manifest V3
- AI Integration: Chrome AI API + Anthropic Claude
- Recording: Custom event capture system
- Storage: IndexedDB with compression
- Build: Vite + CRXJS plugin
We're just getting started! Here's what we're brainstorming for the future:
- One-Click Automation: Convert recorded tasks into executable browser agents
- Playwright/Selenium Export: Generate ready-to-run automation scripts
- Voice-Activated Tasks: "Hey, book my usual flight" → automated execution
- Smart Replay: Re-run tasks with dynamic adaptation to UI changes
- Smart Notifications: "You usually check email after this task"
- Auto-Form Fill: Learn your patterns and pre-populate forms
- Workflow Shortcuts: One-click execution of multi-step processes
- Context-Aware Suggestions: "Based on your current task, you might want to..."
- Multi-Task Learning: Understand relationships between different workflows
- Anomaly Detection: "This task is taking longer than usual"
- Predictive Actions: Suggest next steps before you think of them
- Cross-Site Intelligence: Connect workflows across different websites
- API Webhooks: Trigger external services based on task completion
- Slack/Teams Bots: Share workflow insights with your team
- Analytics Dashboards: Track productivity patterns over time
- Browser Extension Bridge: Connect with other productivity tools
- Visual Workflow Builder: Drag-and-drop task automation
- Mobile Companion: Analyze mobile web workflows
- Team Collaboration: Share and compare workflows with colleagues
- Custom Roasting Themes: Because regular roasting isn't enough 🔥
- Compliance Monitoring: Ensure workflows follow company policies
- Training Mode: Generate onboarding materials from expert workflows
- Performance Metrics: Team productivity insights and optimization
- Security Auditing: Track sensitive data interactions
🚀 Want to help build the future of browser intelligence?
We're looking for contributors interested in:
- AI/ML for workflow prediction
- Browser automation frameworks
- Chrome extension APIs
- UX design for complex workflows
- Developer tooling and integrations
Drop us a line in the issues if any of these directions excite you!
[Add screenshots here of the side panel, analysis results, roasting mode, etc.]
Want to make this even better?
- Fork the repo
- Create a feature branch:
git checkout -b feature/amazing-feature - Commit changes:
git commit -m 'Add amazing feature' - Push to branch:
git push origin feature/amazing-feature - Open a Pull Request
MIT License - see LICENSE file for details.
- Built with Chrome's experimental AI APIs
- Powered by Anthropic's Claude for advanced analysis
Made with ❤️ for anyone tired of explaining what they just did on a website
Star ⭐ this repo if it helped you understand your own browsing habits!