Amicompat

Inspiration

Modern web development moves fast, and ensuring consistent behavior across browsers has become increasingly tricky. We wanted to create a simple yet powerful tool that helps developers understand how compatible their code really is with today’s web standards, without constantly checking MDN, Can I Use, or guessing which polyfills to add.

AmICompat was born from the need for an automatic, local, and reliable way to analyze real-world browser compatibility for web projects.

What it does

AmICompat analyzes your CSS and HTML files to detect modern features that may not yet be widely supported by all browsers.

It then produces a comprehensive compatibility report (CLI or JSON export) that includes:

Detected modern features
Their Baseline support status
Browser versions and compatibility data
Recommendations for improving or adapting your code

In short: AmICompat tells you if your web project is ready for the real world.

How we built it

The project is built with TypeScript and powered by a hybrid detection architecture combining:

ESLint (with @eslint/css and @html-eslint) for robust feature detection
web-features and compute-baseline for browser support enrichment
Zod for type-safe schema validation
MCP (Model Context Protocol) integration to make AmICompat usable from within AI coding environments and editors such as VS Code, Cursor, and Claude Code

This design ensures AmICompat runs fast, and fits naturally into modern development or AI-assisted workflows.

Challenges we ran into

Bridging ESLint detection with live compute-baseline data effectively without performance issues
Handling the wide variety of modern CSS and HTML syntaxes
Ensuring data consistency across MCP tool schemas
Designing developer and AI-friendly reports that are both detailed and readable

Accomplishments that we're proud of

Successfully integrating ESLint Baseline validation directly, without custom parsing
Delivering a stable first release with full test coverage
Enabling native AI integration via the MCP protocol, usable instantly with npx
Making the tool fully open source and accessible to the community

What we learned

The MCP protocol is a powerful bridge between developer tools and LLM environments
ESLint is far more flexible and powerful than a simple linter, perfect for static analysis use cases
Clear, actionable compatibility visualization helps developers fix issues faster