Back to Lab

RemoteCtrl

Browser automation with AI agents

An open-source Electron project for turning prompts into browser actions. RemoteCtrl brings an AI agent loop, Playwright interactions, and visual feedback into a desktop app.

RemoteCtrl illustration

Inside the project

An agent loop

Break browser tasks into steps and turn each step into an executable action.

Browser interactions

Use Playwright to navigate, click, type, scroll, and extract page content.

Visual feedback

Cursor overlays show where the agent is clicking and interacting with a page.

Screen streaming

WebRTC screen capture lets you follow the browser session remotely.

Explore the code

The repository contains the app and setup instructions. It's a place to explore browser agents, page extraction, and automation workflows.