firefox-devtools-mcp-automation

What problem does it solve?

This Skill eliminates manual browser testing and web exploration by enabling AI-driven control of Firefox for repeatable inspection, interaction, and data capture.

Core Features & Use Cases

• AI-assisted browser automation with MCP: Navigate pages, manage tabs, and perform real interactions (click, fill, hover, drag/drop, upload) without relying on brittle CSS selectors.
• Visual and UID-based targeting: Take snapshots that assign UIDs to interactive elements, then act precisely on those UIDs for stable workflows across layout changes.
• Observability for debugging and scraping: Capture screenshots, monitor network requests/responses, and read console messages to validate behavior and extract results.
• Advanced support for privileged control: Optionally use privileged contexts to manage extensions and set Firefox preferences when needed.
• Use case: Build an end-to-end login-and-scrape flow that navigates to a page, fills a form using snapshot UIDs, captures a results screenshot, and extracts relevant API responses from network logs.

Quick Start

Use the MCP server to open Firefox headlessly, navigate to your target URL, take a snapshot to identify element UIDs, and then interact with the page using those UIDs.