mcp-browser

Automate web navigation by directing the browser to a target URL, capturing an accessibility snapshot to identify element roles and refs, and then executing explicit interaction sequences like clicking or typing. This enforces deterministic workflows for reliable UI testing and debugging.

An accessibility snapshot captures the accessibility tree of a web page, exposing element roles, names, and refs. It maps UI elements for reliable targeting during browser automation, ensuring that clicks, typing, and form interactions hit the correct elements.

Fill forms and simulate keyboard input by capturing an accessibility snapshot first, then using the extracted element refs to target input fields. You execute explicit action sequences to type text, select options, and simulate keyboard events for deterministic form workflows.

Yes, you can inspect console logs to debug issues during automated browser interactions. The Skill captures console and log output alongside accessibility snapshots and UI interactions, allowing you to diagnose problems quickly while executing action sequences.

Yes, taking an accessibility snapshot before interacting with elements is required. The Skill enforces deterministic workflows by requiring snapshots before interactions, ensuring you have explicit element refs for accurate targeting during clicking, typing, and hovering actions.

Accessibility snapshots require explicit element refs for targeting and enforce strict sequential workflows, meaning you cannot interact with elements before capturing a snapshot. This structured approach ensures deterministic web testing but requires navigating pages step by step.