xavi-999/metrognome

README preview

![metrognome — the autonomous performance engineer for React Native](docs/banner.png)

[![CI](https://github.com/xavi-999/metrognome/actions/workflows/ci.yml/badge.svg)](https://github.com/xavi-999/metrognome/actions/workflows/ci.yml)
[![Release](https://img.shields.io/github/v/release/xavi-999/metrognome)](https://github.com/xavi-999/metrognome/releases)
[![npm](https://img.shields.io/npm/v/metrognome)](https://www.npmjs.com/package/metrognome)
[![npm downloads](https://img.shields.io/npm/dm/metrognome)](https://www.npmjs.com/package/metrognome)
[![License: MIT](https://img.shields.io/badge/license-MIT-16a34a)](./LICENSE)
[![Live demo](https://img.shields.io/badge/live_demo-xavi--999.github.io-16a34a)](https://xavi-999.github.io/metrognome/)

> **The autonomous performance engineer for React Native.** `/metrognome`: one command turns scattered RN performance tooling into a single, scientific loop — **propose → measure → keep/revert** — that ships gains it can prove.

For any React Native developer chasing TTI, memory leaks, jank, bundle size, or re-renders: one `/metrognome` call researches, optimizes, and verifies autonomously — and every gain it reports is one it measured.
One input, zero human interaction, flabbergasting performance improvements.

Works with **Claude Code, OpenAI Codex CLI, Cursor, Gemini CLI, and GitHub Copilot CLI.**

Inspired by OpenAI co-founder Andrej Karpathy's `/autoresearch` [skill](https://github.com/karpathy/autoresearch). Fully adapted to React Native.

## What `/metrognome` solves

View full repository on GitHub →