Browse, extract, and automate rendered web workflows.
browser automation
Automates browser navigation, interaction, rendered-content extraction, network capture, screenshots, forms, and parallel sessions.
When to use it
Use for browser-act requests and browser work involving rendered pages, JavaScript, authenticated sessions, forms, screenshots, or browser management.
Give it a browser task; it fetches or extracts rendered content and performs requested browser interactions.
This skill
CLI data directory
Stores and deletes browser profiles (irreversible)
web forms
Submits forms and uploads files
verification-assistance API
Sends captcha challenge images
web pages
Reads rendered web pages
Requires Python 3.12 or newer.
Requires the uv package manager.
The browser-act CLI must be available to retrieve the version-matched workflow.
A local Chrome instance is required only for the chrome-direct browser type.