Record your browser actions and generate automation scripts that understand what you're trying to do. Run with your own credentials on your own machine — no cloud, no data leaving your browser. When something breaks, it fixes itself and keeps going. Full access to your local files for real-world workflows.
Record browser interactions and get automation code. This guide walks you through installation, capturing your first interaction, and running automated workflows.
Download and install the InverseUI Chrome extension to start capturing UI interactions:
chrome://extensionsOnce installed, you'll see the InverseUI icon in your browser toolbar.
Before recording, sign in to your InverseUI account:
Let's capture your first UI interaction and generate automation code.
Click the InverseUI extension icon in your toolbar and select Start Recording. You'll see a "rec" badge indicating recording is active.
Now perform any actions you want to automate: - Click buttons and links - Fill out forms - Navigate between pages - Upload files - Use keyboard shortcuts
When you're done, click the extension icon again and select Stop Recording.
InverseUI automatically: - Uploads your recording to the cloud - Generates Playwright automation code - Opens your recording in the dashboard
Your recording is now available at inverseui.com/track/{track_id}:
| What You Get | Description |
|---|---|
| Recorded Actions | Every click, input, and navigation captured |
| Generated Code | Clean Playwright code with typed parameters |
| Intent Detection | Smart comments explaining what each action does |
| Reusable Functions | Parameters auto-detected for flexible automation |
Download the generated code or copy it directly into your project.
Run your recorded automations locally with automatic error recovery.
Requires Python 3.10+ and Google Chrome.
pip install inverseui
Authenticate via browser, then paste the code into your terminal.
inverseui login
Execute a recorded track by its ID.
inverseui run <track_id>
Automatically retry failing scripts with AI-generated fixes.
inverseui fix <track_id>
Control recording from your own scripts using the injected API:
// Available on any page when extension is installed
const api = window.INVERSEUI_EXTENSION_API;
await api.startRecording();
// ... perform actions ...
await api.stopRecording();
const actions = await api.getActions();
For the complete API reference, see Developer Guides.
AGPL-3.0 — If you modify or use this code, you must publish your changes under the same license.
$ claude mcp add InverseUI-Recorder \
-- python -m otcore.mcp_server <graph>