browser

The Browser skill provides automated control over a Chrome session within the Factory CLI environment via Stagehand. It allows users to interact with websites programmatically, perform actions, extract structured data, and capture screenshots. This skill is designed to streamline repetitive web interactions, automate data collection, and provide a reliable method for recording and verifying web-based workflows. By integrating live browser automation directly into the Factory session, it eliminates the need for manual browsing and ensures consistent results across tasks.

Key features include the ability to navigate to URLs, execute natural-language-driven actions, extract structured data based on a specified schema, observe suggested next steps, and capture screenshots of the current viewport. Screenshots are automatically saved to a dedicated directory, and the skill provides detailed summaries of completed tasks along with follow-up recommendations. The skill also ensures proper session management by requiring Chrome to run via Stagehand and by safely closing sessions to prevent lingering processes.

This skill is particularly useful for developers, QA engineers, and data analysts who need to automate web interactions for testing, data gathering, or reporting. It supports complex flows that would otherwise require manual intervention, making it ideal for repetitive web tasks, automated monitoring of sites, and structured data extraction. Users who require precise control over a Chrome session, want to integrate browser actions into automated pipelines, or need consistent screenshots for documentation or analysis will find this skill highly valuable.