Flutter 自动化测试工具 是 AI Skill Hub 本期精选MCP工具之一。综合评分 7.5 分,整体质量较高。我们推荐使用将其纳入你的 AI 工具库,帮助提升工作效率。
The first MCP server for autonomous Flutter testing on real iPhones and Android,支持Flutter自动化测试。
Flutter 自动化测试工具 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
The first MCP server for autonomous Flutter testing on real iPhones and Android,支持Flutter自动化测试。
Flutter 自动化测试工具 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/michal-giza/flutter-dev-agents
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"flutter--------": {
"command": "npx",
"args": ["-y", "flutter-dev-agents"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 Flutter 自动化测试工具 执行以下任务... Claude: [自动调用 Flutter 自动化测试工具 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"flutter________": {
"command": "npx",
"args": ["-y", "flutter-dev-agents"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
The first MCP server that lets autonomous agents build, deploy and test Flutter apps on real iPhones and Android devices.
137 tools across Android (uiautomator2 + adb), iOS (WebDriverAgent + pymobiledevice3), Flutter (Patrol + flutter run --machine), and a 7-vertical opinionated audit suite for shipping with confidence. Works with Claude Desktop, Claude Code, Cursor, or any MCP-aware host. Composes with Google's official Dart/Flutter MCP and Maestro MCP — see the stack.
→ First 15 minutes · The Stack · Senior-tester discipline · Comparison vs other MCPs · FAQ · Configuration · Operational gotchas · Tools by category · Architecture
The Maestro composition release. We now sit explicitly on top of Maestro (mobile.dev's flow-based mobile test framework, whose MCP launched Feb 2026) — auditing what their flows produce rather than competing with them. Same posture for Google's official Dart/Flutter MCP.
audit_maestro_flow — lint Maestro YAML flows against 12 senior-tester rules (hardcoded locale strings, vacuous assertions, sleep_in_flow, missing failure paths, …)ingest_maestro_report — parse Maestro execution reports (JUnit XML + Maestro JSON), surface flake / regression signalsaudit_release_readiness — extended with a 6th test_execution domain (opt-in via maestro_report_path); failed flows propagate to verdict=blockSee the stack for how the 4 MCPs compose end-to-end, and the comparison memo for the full landscape analysis.
Previous milestones: - v0.3.1 — calibration patches from 3-project field test, signal:noise ~96% - v0.3.0 — the 7-vertical audit suite (seniority + security + i18n + supply chain + a11y + test-quality + composite gate) + senior-tester loop (design_test_plan + audit_test_quality) - v0.2.x — initial PyPI release, multi-device locking, Patrol integration, AR/vision
---
| Path | What |
|---|---|
[packages/phone-controll/](packages/phone-controll/) | The flagship MCP. **137 tools** spanning device control, build/install/launch, Patrol-driven Flutter UI tests, AR/Vision, declarative YAML test plans, cross-session device locking, the **7-vertical audit suite** (seniority/security/i18n/dependencies/a11y/test-quality + composite), the **senior-tester loop** (design_test_plan + audit_test_quality), and **Maestro composition** (audit_maestro_flow + ingest_maestro_report). |
packages/<future>/ | Future MCPs slot in here using the same shape (see [docs/adding_an_mcp.md](docs/adding_an_mcp.md)). |
[examples/templates/](examples/templates/) | Shared YAML test-plan templates (smoke, ump-decline, ar-anchor, flutter-test-smoke). |
[examples/agent_loop.py](examples/agent_loop.py) | Reference autonomous Plan→Build→Test→Verify loop using any OpenAI-compat local LLM. |
[skills/](skills/) | Symlinks to the Claude Code skills that ship with these MCPs. |
[scripts/](scripts/) | Fresh-laptop installer, doctor, and ops scripts. |
[docs/](docs/) | Architecture, framework-extension recipe, MCP-extension recipe. |
See packages/phone-controll/README.md for the full list. Briefly:
adb (brew install --cask android-platform-tools)pymobiledevice3 remote tunneld running for developer-tier servicesflutter on PATH; for Patrol: dart pub global activate patrol_cli[ar] extra installs OpenCV[http] extra installs FastAPI + uvicornRun check_environment from any Claude Code session — it returns a structured doctor report with concrete fix commands for any red items.
```bash git clone <this repo> ~/Desktop/flutter-dev-agents cd ~/Desktop/flutter-dev-agents/packages/phone-controll uv venv --python 3.11 uv pip install -e ".[dev,ar,http]" pytest # full unit suite, no toolchain needed
A typical day on the factory laptop:
Claude #1 in checkaiapp/
→ open_project_in_ide("checkaiapp") # spawns its own VS Code window
→ select_device(R3CYA05CHXB) # acquires the lock on the Galaxy
→ start_debug_session(project_path=...) # `flutter run --machine`, returns vm_service_uri
→ ...edit code, restart_debug_session, read_debug_log, repeat...
→ run_patrol_test (or run_test_plan with dev_iteration.yaml)
→ stop_debug_session, release_device, close_ide_window
Claude #2 in another_app/ → emulator-5554, its own VS Code, its own debug
Claude #3 in third_app/ → iPhone simulator UDID, its own VS Code, its own debug
Three independent debug sessions, three IDE windows, three locked devices, no collisions. The HTTP adapter exposes both the unified /tools/* surface and a focused /dev-session/* sub-router for agents that only care about the dev-iteration loop.
See examples/templates/dev_iteration.yaml for a runnable plan template; docs/ios_setup.md for the iPhone prerequisites (Developer Mode, DDI, tunneld, WebDriverAgent).
该工具提供了一个开源的MCP服务器,支持Flutter自动化测试,虽然star数较少,但仍然是一个有价值的工具,值得关注。
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ Apache 2.0 — 宽松开源协议,可商用,需保留版权声明和 NOTICE 文件,含专利授权条款。
经综合评估,Flutter 自动化测试工具 在MCP工具赛道中表现稳健,质量良好。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | flutter-dev-agents |
| 原始描述 | 开源MCP工具:The first MCP server for autonomous Flutter testing on real iPhones and Android 。⭐5 · Python |
| Topics | mcpagentai-agentsandroidanthropicappiumpython |
| GitHub | https://github.com/michal-giza/flutter-dev-agents |
| License | Apache-2.0 |
| 语言 | Python |
收录时间:2026-06-06 · 更新时间:2026-06-11 · License:Apache-2.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端