Maestro

Why we included this project

Maestro writes and runs end-to-end UI tests for Android, iOS, and web from plain YAML flows. You describe interactions as commands like launchApp, tapOn, and assertVisible, and the interpreted engine handles the flakiness tolerance and smart waiting that usually eat up a testing budget. The part that matters for AI-assisted work is the MCP server bundled with the CLI: coding agents such as Claude Code or Cursor can inspect a live emulator's view hierarchy, write a flow, run it, read the failure, and self-correct until the test passes. The flows stay human-readable YAML you own and can run in CI long after the chat ends, so you get agent-assisted testing without giving up reproducibility.