web-clone

Cloning a webpage with visual and functional parity involves using Playwright MCP to extract the live DOM, generating code via an LLM, and iteratively verifying the output. This Skill produces a runnable scaffold that mirrors layout, interactions, and responsive behavior.

Yes, a Playwright MCP server is required to reproduce interactive functionality. The workflow depends on browser tools such as browser_navigate, browser_snapshot, and browser_take_screenshot to extract the live page and verify the generated clone scaffold.

The best way to generate a runnable clone scaffold from a URL is using LLM-driven code generation guided by live page extraction. This workflow captures the target site's appearance and interactions, outputting a project scaffold for rapid prototyping.

Visual verification works by taking screenshots of the target and cloned pages, then comparing them to produce a visual verdict score. This score guides iterative improvements to the generated code until visual fidelity is achieved.

Yes, you can clone sites with complex forms and modals. The workflow preserves key interactive functionality like navigation and form inputs by evaluating the live page's behavior and replicating it in the generated scaffold.

Limitations include dependency on the Playwright MCP server's availability and the LLM's ability to interpret complex DOM structures. Highly dynamic content or authenticated areas may require additional manual configuration to achieve functional parity.