Use when writing acceptance criteria for a task - express each as an observable Given/When/Then that QA can execute, including negative cases
日本語の概要は準備中です。原文の説明を表示しています。
Use when starting a mobile task that could be Flutter or native - decide from the existing app, platform-specific needs, and the analiz task
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Flutter is the default for cross-platform mobile; native (SwiftUI / Jetpack Compose) is chosen when a task needs deep platform integration or the app is already native. Decide at the start — you cannot half-migrate an app mid-task.
Core principle: The existing app's stack wins by default. Only greenfield apps or the analiz task's direction open the choice.
digraph m {
"App already single-stack?" [shape=diamond];
"Follow the app's stack" [shape=box];
"Needs deep platform integration?" [shape=diamond];
"Native (SwiftUI / Compose)" [shape=box];
"Flutter (default)" [shape=box];
"App already single-stack?" -> "Follow the app's stack" [label="yes"];
"App already single-stack?" -> "Needs deep platform integration?" [label="greenfield"];
"Needs deep platform integration?" -> "Native (SwiftUI / Compose)" [label="yes"];
"Needs deep platform integration?" -> "Flutter (default)" [label="no"];
}
Never introduce a second UI stack into a single-stack app. A Flutter app that "needs one native screen" embeds it via a platform view (PlatformView/AndroidView/UiKitView) or talks to it through a MethodChannel/Pigeon — it does not gain a parallel SwiftUI module on your initiative. Going the other way (a native app needing a Flutter module) is Flutter add-to-app, not a rewrite. Flag genuine cross-stack needs in a task comment for the architect.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Use when writing acceptance criteria for a task - express each as an observable Given/When/Then that QA can execute, including negative cases
日本語の概要は準備中です。原文の説明を表示しています。
Use when the diff adds or changes an endpoint, resolver, RPC, job or query that takes an object id, a role check, a request binding or a tenant filter - BOLA/IDOR, function-level authorization, mass assignment and tenant scoping
日本語の概要は準備中です。原文の説明を表示しています。
Use on every UI change - semantic HTML, labels for controls, keyboard-navigable dialogs/menus, visible focus, and never color as the only signal
日本語の概要は準備中です。原文の説明を表示しています。
Use when a task changes any screen, form, dialog, menu or control - Lighthouse/axe scan of the changed screens, a keyboard walk, and the thresholds that fail a task
日本語の概要は準備中です。原文の説明を表示しています。
How to work a task returned with review, QA or UAT findings. Use when a task is in need_revision or PR review comments are in your context.
日本語の概要は準備中です。原文の説明を表示しています。
Use when deciding whether a request needs an analiz task before implementation - the conditions that require the architect's analysis versus going straight to implementation
日本語の概要は準備中です。原文の説明を表示しています。