Browser Automation

Drive a browser with snapshot refs to navigate, click, fill forms and screenshot

Automate browsers with Playwright or Puppeteer using reliable selectors and waits

Build agents that view screens, click and type with safe sandboxing

Run website tasks in Dot browser identities and pick the right sign-in profile

Drive native desktop apps through accessibility snapshots and element tokens

Create and clean up disposable Linux or macOS sandboxes to run code and browse

Run commands and show desktops inside remote or local Spaces you can watch

Test buttons, fill forms and run visual QA by driving a GUI with the cua CLI

Build a bounded computer-use loop that observes, acts and verifies each step

Browse websites to read pages, fill forms, take screenshots and extract data

Write OpenCLI site adapters from observed browser behavior and API evidence

Drive Chrome with OpenCLI to inspect pages, use logged-in sites and capture API calls

Find and run OpenCLI commands for websites and Chrome and read their results

Capture and automate a real browser from the terminal with Playwright

Automate websites with the agent-browser CLI for forms, clicks and screenshots

Run the Obscura headless browser for page loading, screenshots and CDP automation

Automate browser tasks, form filling and data extraction from web pages

Drive Google NotebookLM from the browser to query notebooks and add sources

Drive native GUI apps on macOS, Windows and Linux via accessibility snapshots

Open sites in cmux, wait for page state and extract data without stealing focus

Drive real macOS apps from a cmux session with screenshots, clicks and typing

Read files, run commands and drive the desktop of a connected laptop or server

Operate websites in a browser that shares your logged-in state with agents

Drive a headless browser from the CLI using accessibility snapshots and element refs

Drive real Chrome or Brave and macOS apps for visual deploy checks and screenshots

Control browser tabs, clicks, forms, downloads and screenshots on web pages

Build, install and audit Chrome extensions inside the Agent Zero browser

Fill and submit web forms with uploads, dropdowns and rich text fields

Control the host desktop with screenshots, clicks and typing in a beta tool

Guide Linux desktop control with AT-SPI targeting and screenshot checks

Guide macOS desktop control with Accessibility tree targeting and window actions

Guide Windows desktop control with UI Automation targeting and session caveats

Operate a sandboxed Linux desktop and its GUI apps in a Docker or Xpra session

Control your real logged-in Chrome over CDP for clicks, typing and navigation

Automate browser interactions with Playwright patterns for autonomous web tasks

Act on websites from the shell with stealth browsers and saved per-person sign-ins

Run web tasks in a persistent signed-in browser for reading, forms and downloads

Script common web login flows with Playwright for tests and browser tasks

Open, inspect, click through and screenshot web pages in a real browser

Drive the embedded browser with the cheapest reliable interaction loop

Drive the user's running Chrome from the terminal with the hive-browser CLI

Automate LinkedIn posting and messaging in the browser, avoiding known page traps

Automate browser tests, form filling, screenshots and data extraction with Playwright

Automate browser tasks with PinchTab: navigate, fill forms, scrape and export PDFs

Automate your logged-in Chromium browser to read pages, fill forms and scrape data

Give an agent full web access for search, page fetching and logged-in browsing

Automate your existing Chrome browser over CDP while keeping your logins

Handle login walls, CAPTCHAs, lazy loading and paywalls in the managed browser

Click and type in macOS apps on a local or remote host through computer-use tools