aboul3ata/lazyweb-skill

What it solves

Lazyweb provides AI coding agents with empirical design evidence—such as real product screenshots, UI references, and design research workflows—to replace "vibes-based" design with evidence-backed interfaces. It bridges the gap between an agent's ability to write code and its lack of real-world UI/UX benchmarks.

How it works

It functions as a skill pack and a hosted Model Context Protocol (MCP) server. Once installed, it integrates into coding clients (like Cursor, Claude Code, or Codex) by providing a set of slash commands and MCP tools. These tools allow agents to search for real product screens, study multi-step user journeys (onboarding, checkout), and find documented growth experiments to guide their design and conversion decisions.

Who it’s for

Developers and product designers using AI coding agents who want to ensure their UI/UX is based on proven industry patterns and real-world product evidence rather than generic AI guesses.

Highlights

• Evidence-Based Design: Access to real product screenshots and UI patterns to guide conversion and design decisions.
• Growth Research: Tools to score a website's growth readiness and generate prioritized design improvement reports.
• Experiment Discovery: Ability to search for real-world pricing, paywall, and monetization experiments.
• Broad Integration: Supports a wide range of agentic IDEs and clients including Cursor, Claude Code, Codex, and others via MCP.