artemis

Google's Artemis drives real Android phones over MCP and keeps the logs as it goes. The claim is 99%+ on AndroidWorld. Mobile is where the agent pitch usually gets vague, so that claim is doing a lot of work. google/artemis #android #ai

Test Automation with agent-device (Part 1) --- title: "Test Automation with agent-device (Part 1)" published: true description: "Why agent-device for Flutter app testing, how to install it for Claude Code, the basic command loop, and Flutter gotchas." tags: claude, ai, flutter, testing canonical_url: krenin88.github.io --- agent-device is an open-source tool from Callstack that lets an AI agent drive iOS and Android apps and check what is on screen. This post covers why it was picked for testing a Flutter app, how to install it for Claude Code, the basic command loop, and what to watch out for with Flutter. It is based on version 0.21.31 (2026-10-24). Why agent-device Flutter ships Android and iOS apps from one codebase, so the device-testing tool has to cover iOS too. Google Artemis, the first candidate, is a natural-language testing agent, but it supports Android only; iOS is still just a roadmap item. The other candidates looked like this: - mobile-mcp exposes taps, screenshots, and the accessibility tree on iOS and Android over MCP (the protocol an AI uses to call external tools). It has no sessions, replay, or evidence storage, so a verified flow is hard to hand off to CI. - Maestro is an E2E framework that pins flows in YAML, and it has an MCP server. It suits regression runs of fixed flows better than an agent exploring the screen. - Arbigent is an AI agent built on top of Maestro. Its author rates it 1/5 for speed and 3/5 for reliability, and its YAML format was still changing. - Commercial services (Drizz, Minitap, and others) bundle a real-device farm with a natural-language agent, but the only public material was vendor material, so cost and quality could not be checked. agent-device handles iOS and Android simulators, emulators, and real devices with one tool, and its docs state Flutter support explicitly. It ships both a CLI and an MCP server, can save a verified flow as an .ad script to replay later, and can export to Maestro YAML. It is MIT-licensed. Installation Install the CLI globally and check the environment. npm i -g agent-device@latest agent-device doctor doctor checks connected devices, toolchains, and dev-server status. In this setup it passed with only warnings for unused platforms (HarmonyOS, Vega). Register it with Claude Code as an MCP server. With -s user, the same tool is available in every project. claude mcp add -s user agent-device -- agent-device mcp The agent-device repository includes three agent skills: agent-device, ios-simulator, and android-emulator. The official instruction is npx skills add callstack/agent-device, but Claude Code's auto mode safety check blocked it. In that case, clone the repository and copy the three folders under skills/ into ~/.claude/skills/. To undo everything, run claude mcp remove agent-device -s user and npm uninstall -g agent-device, then delete the three skill folders. The basic loop The basic loop is: prepare a device, open the app, act, verify, close. Pick the device with --platform and --device. agent-device boot --platform android --device "<emulator name>" --headless agent-device install /absolute/path/app-debug.apk --platform android --device "<emulator name>" agent-device open <bundle-id> --platform android --device "<emulator name>" --foreground agent-device press @e5 --settle agent-device fill @e11 "text" --settle agent-device wait text "<expected text>" agent-device screenshot /absolute/path/home.png agent-device close open reads the screen and returns a ref such as @e5 for each element, like buttons and text fields. With --settle, press and fill wait until the screen goes quiet and return only what changed, so the agent can pick its next target without rereading the whole screen. Three things commonly go wrong here: - Refs go stale when the screen changes. Pick a new ref from the diff returned by each action. - A screenshot is not verification. Verify with wait text or wait <selector>, and keep screenshots as evidence only. - Give install an absolute path. A relative path resolves against the agent-device daemon's working directory and fails with "App binary not found". For a Flutter app, build the iOS simulator app with flutter build ios --simulator --debug and the Android app with flutter build apk --debug. Flutter gotchas agent-device reads the screen through the accessibility tree. In Flutter, that means only widgets with text show up by default; icon buttons and containers do not. Add Semantics(identifier:) or semanticLabel to those widgets so a selector can find them. Flutter's Key does not appear in the accessibility tree. Labels in the accessibility tree follow the language the device is set to, so the same app reads with different label text depending on that setting. A text-based selector changes with the device language, which makes identifier the more stable choice. On iOS, AdMob test banners show up as platform_view, webview, and link nodes. Selectors in replay scripts should avoid those nodes. References - agent-device repository: <callstack/agent-device> - agent-device agent setup docs: <oss.callstack.com> - Google Artemis: <google/artemis> - Maestro Flutter support docs: <docs.maestro.dev>

google / artemis ARTEMIS turns natural-language instructions into reliable Android automation. It automates end-to-end workflows, captures logs, and integrates seamlessly with AI coding assistants such as Antigravity, Codex, and Claude Code. It also achieves 99%+… google/artemis

Top 5 open source do GitHub | Vol. 11 #shorts Os projetos #5 a #1 do Top 15 open source do GitHub, Vol. 11, 10 segundos para cada um. #5 Rustrak: rustrak/rustrak #4 Codewhale: codewhale-hq/Codewhale #3 ESP32 iPhone Agent: onehumanbeing/esp32-iphone-agent #2 Artemis: google/artemis #1 Mission Control: builderz-labs/mission-control Vídeo completo com os 15: youtu.be Oferecimento: Hallen Tech, hallentech.com #shorts #opensource #github

Mais 15 projetos open source do GitHub que valem o seu tempo | Vol. 11 Mais 15 projetos open source do GitHub que valem o seu tempo, em contagem regressiva e com 10 segundos para cada um: de uma placa de US$ 7 que deixa a IA usar um iPhone a um painel para operar agentes de IA, passando por avatar animado, casa 3D no Home Assistant e o fim das travadas de shader nos jogos. 📺 Décimo primeiro volume da série Open source do GitHub. Escolhidos pela análise da lista inteira (estrelas, frequência de commits, último commit, README, documentação, capturas e site), conferida em 21/10/2026. Qual desses você vai testar primeiro? Conta nos comentários! ⏱️ CAPÍTULOS E LINKS 0:00 Abertura 0:10 #15 tapflow: simuladores de iPhone e emuladores Android no navegador, no seu servidor (1,2 mil estrelas) jo-duchan/tapflow 0:20 #14 Mesh Avatar Studio: transforma uma ilustração num avatar 2D animado que pisca, fala e mexe a cabeça shinshin86/mesh-avatar-studio 0:30 #13 openGym: treinos guiados, séries e peso no celular, com os dados no seu servidor (14,5 mil estrelas) DuarteSantos8/openGym 0:40 #12 SCSKiller: compila os shaders antes do jogo para acabar com as travadas no PC (Windows) (3,3 mil estrelas) BlueHeisenberg/SCSKiller 0:50 #11 OpenFilm: o seu agente de código faz o vídeo como páginas web e você edita no navegador (1,9 mil estrelas) openfilm/openfilm 1:00 #10 Humanizer: skill que faz texto de IA soar escrito por gente, com base no guia da Wikipédia (59,3 mil estrelas) blader/humanizer 1:10 #9 NeonPlan 3D: planta 3D neon da sua casa dentro do Home Assistant, com luzes e câmeras ao vivo Mastershort/neonplan3d 1:20 #8 MangoDisk: limpeza de disco, análise de armazenamento e privacidade para macOS, Windows e Linux (4,5 mil estrelas) harry0703/MangoDisk 1:30 #7 OpenUsage: uso das assinaturas de IA de código na barra de menu do Mac (4,4 mil estrelas) robinebers/openusage 1:40 #6 Finder Menu: as opções que faltam no clique direito do Finder: novo arquivo, terminal aqui, copiar caminho nikgrbn/finder-menu 1:50 #5 Rustrak: rastreador de erros self-hosted compatível com os SDKs do Sentry rustrak/rustrak 2:00 #4 Codewhale: agente de código no terminal que funciona com mais de 40 provedores de modelos (41,1 mil estrelas) codewhale-hq/Codewhale 2:10 #3 ESP32 iPhone Agent: uma placa ESP32 de US$ 7 deixa um modelo de IA usar um iPhone de verdade, sem computador (3,5 mil estrelas) onehumanbeing/esp32-iphone-agent 2:20 #2 Artemis: IA e testes usando celulares Android reais, com MCP para Claude Code e outros (13,6 mil estrelas) google/artemis 2:30 Oferecimento Hallen Tech 2:44 #1 Mission Control: painel self-hosted para operar agentes de IA: tarefas, execuções e gastos (6,4 mil estrelas) builderz-labs/mission-control 2:54 Encerramento 🧡 Oferecimento: Hallen Tech, liderança de tecnologia sob demanda, software, cloud e segurança para a sua empresa. Site: hallentech.com E-mail: contato@hallentech.com Instagram: instagram.com LinkedIn da Hallen Tech: linkedin.com LinkedIn do Eddwan: linkedin.com Se curtiu, deixa o like, segue o canal e ativa as notificações para a próxima lista. #opensource #github #programacao #ferramentas

Google quietly open-sourced an agent that taps Android phones for you. I assumed another demo repo — then I read the README. ARTEMIS. Apache 2.0, Python 3.12+ google/artemis It runs over MCP, so Claude Code, Codex or Cursor drives a real device or emulator: clicks, swipes, text input across apps. Screenshots and Logcat come back with the report. Two profiles, and the gap between them is the whole story: Flash — reactive loop, ~3-5s per step, no plan, no safety net Pro — Planner + Operator + Checker, ~15-40s per step, every action passes a safety net before it fires 99%+ on AndroidWorld, 100+ multi-step tasks. Android only — iOS is on the roadmap, not shipped. The honest part: this replaces the tapping, not the judgment. You still decide what's worth testing. #AI #OpenSource

ARTEMIS: Android-автоматизация на нейросетях #shorts ARTEMIS: Google учит ИИ управлять Android как человек Представьте: вы пишете «открой карты, построй маршрут, запусти Coldplay», и телефон сам всё делает. Google выкатил ARTEMIS - open-source фреймворк, который превращает естественный язык в надёжную автоматизацию Android. ⚡ Главные фичи: • 99%+ успеха на AndroidWorld (100+ задач) • MCP-интеграция с Claude Code, Antigravity, Codex • Захват логов и скриншотов для отладки 🛠 Быстрый запуск: git clone google/artemis &amp;&amp; cd artemis 💡 Где пригодится: автотесты мобильных приложений без ручного тыканья. 🔗 Официальный репозиторий #софт #opensource 👉 Telegram: @the_sintez #shorts #нейросети #ai #софт #технологии

Google Artemis: Plain Words Drive a Real Android Phone Artemis is an open source Android automation agent from Google. You type a goal in plain words, and it taps, swipes and types on a real phone or emulator. It ships a command line tool, a local web console, a Python SDK and an MCP server that AI coding assistants can call. Apache 2.0. google/artemis

OpenAI 互換APIでも動くっぽいのでとりあえずGemma4でAndroidアプリの自動テストしてみる。 google/artemis

ARTEMIS: Let AI assistants and test suites use real phones like a human (12k stars) What it does All text below is quoted directly from the project README. >ARTEMIS turns natural-language instructions into reliable Android automation. It automates end-to-end workflows, captures logs, and integrates seamlessly with AI coding assistants such as Antigravity, Codex, and Claude Code. Key Highlights: - Cross-App Automation: Executes testing workflows and everyday tasks on Android from natural language instructions. - Multimodal Targeting: Uses element indices when available, with coordinate and visual locating fallbacks for custom interfaces. - IDE Diagnostics: Model Context Protocol (MCP) integration lets Antigravity, Claude Code, and Windsurf drive test devices and collect Logcat output and screenshots. - Flash Execution: A reactive observe-and-act loop with asynchronous history summaries, typically 3-5s per step. - Pro Exploration: Checks targets before individual actions and returns blocked actions to the Operator for recovery. Supports long-running exploratory and stability tests. - AndroidWorld Results: 99%+ task completion on Google Research's AndroidWorld benchmark (100+ multi-step tasks). >Ensure an Android device (with USB Debugging enabled) or emulator is connected. The one-click startup script will automatically: Install System Toolchains: Detect and auto-install ADB, scrcpy, FFmpeg, and Python (uv) dependencies. Repository google/artemis Repository stats |Metric|Value| |:-|:-| |Stars|12,067| |Forks|1,315| |Last release|None| |Languages|Python 77.5%, TypeScript 7.7%, HTML 7.2%| |Contributors|6| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|None| |AI co-authored commits|None| Main contributor |Field|Value| |:-|:-| |Name|somew1nd| |GitHub|@somew1nd| |Commits to this repo|89 (top contributor)| |Company / Location|Google| |Followers|44| |Public repos|11| |On GitHub since|2022-12-08|