AI skills focused on Flutter development — prompts, workflows, and tools designed to help AI agents build, audit, refactor, and maintain Flutter applications efficiently.
This repository contains two categories of AI-powered Flutter skills:
Builder skills — guide an AI agent to implement production-grade Flutter code from scratch, following clean architecture, advanced state management, and industry best practices.
Auditor skills — guide an AI agent to perform structured, evidence-based analysis of an existing Flutter codebase, producing a prioritized report with findings and recommendations. Each auditor skill has a focused scope and does not overlap with the others.
A senior-level implementation skill for building enterprise-grade Flutter applications.
Use for: building new screens or features, setting up architecture from scratch, implementing state management with BLoC/Cubit, configuring Firebase, integrating Stitch/Figma designs into Flutter code.
References: architecture · bloc-patterns · dependency-injection · firebase · navigation · performance · stitch-to-flutter
Audits the overall architectural quality of a Flutter project.
Use for: evaluating layer separation, feature modularization, dependency boundaries, state management strategy, and long-term scalability. Produces an Architecture Score (1–10) with a maturity level and prioritized recommendations.
Does not cover: runtime performance, widget rebuild cost, state management internals, design system compliance, testability.
References: architecture · bloc-patterns · dependency-injection · navigation
Audits widget rebuild behavior, render tree efficiency, layout costs, and scroll performance.
Use for: detecting unnecessary rebuilds, expensive build() methods, inefficient list implementations, saveLayer misuse, InheritedWidget propagation issues, Sliver performance problems, and GC allocation pressure.
Does not cover: isolate usage, async efficiency, architecture quality, design system compliance.
References: rebuild-patterns · savelayer-and-compositing · scroll-performance · layout-cost
Audits state management architecture — how state is created, scoped, mutated, and consumed.
Use for: detecting rebuild inefficiencies, inconsistent pattern usage across features, state scoping and lifecycle issues, BLoC/Cubit/Riverpod architectural coupling, unification strategy when multiple solutions are mixed.
Does not cover: widget rendering performance, architecture layer separation, DI quality, testability.
References: state-patterns · scoping-and-lifecycle · rebuild-efficiency · isolation-and-coupling
Audits Dart runtime performance — isolate usage, async efficiency, memoization, caching, and event loop health.
Use for: detecting main-thread blocking work, sequential async calls that should be parallel, missing memoization in expensive computations, cache absence for repeated I/O, scheduleMicrotask misuse, and Timer.periodic event loop saturation.
Does not cover: widget rebuild costs, frame budget, architecture quality, state management.
References: isolate-patterns · async-efficiency · memoization-and-caching · event-loop-health
Audits whether the Flutter UI follows Atomic Design principles and maintains a healthy design system.
Use for: detecting missing atomic hierarchy (atoms/molecules/organisms/templates/pages), import direction violations, duplicated UI primitives, raw hardcoded values instead of design tokens, oversized organisms, and design system scalability gaps.
Does not cover: runtime performance, architecture quality, state management, testability.
References: atomic-hierarchy · token-usage · component-composition · duplication-detection
Audits whether the architecture enables scalable, reliable testing across all test types.
Use for: detecting architectural barriers to unit tests, bloc tests, widget tests, golden tests, and integration tests — DI coupling, hidden dependencies, business logic in widgets, BuildContext in Blocs, non-deterministic widget rendering, missing repository abstractions.
Does not cover: runtime performance, widget rebuild cost, design system quality, architecture layer quality (those belong to other auditors).
References: unit-test-architecture · bloc-test-patterns · widget-test-patterns · golden-test-prerequisites
Audits Flutter CI/CD pipelines on GitHub Actions — structure, caching, test sharding, platform build signing, release automation, and monorepo pipeline design.
Use for: diagnosing slow pipelines, detecting redundant toolchain invocations, configuring test sharding, wiring Android/iOS/Web build signing, gating release automation on quality jobs, evaluating monorepo multi-package pipeline strategies.
Does not cover: Flutter application code quality, test coverage targets, Dart lint rule configuration, self-hosted runner infrastructure.
References: diagnosis · testing · builds · quality · release · gh-actions
| Question | Skill to use |
|---|---|
| "Build a screen / feature / integration" | flutter-scalable-app |
| "Is our architecture scalable?" | architecture-auditor |
| "Why is the UI janky / rebuilding too much?" | widget-performance-analyzer |
| "Is our BLoC/Riverpod usage correct?" | state-management-auditor |
| "Why is our app slow on heavy data operations?" | runtime-performance-auditor |
| "Is our design system consistent?" | atomic-design-system-auditor |
| "Can we write unit / widget / golden tests?" | testability-architecture-auditor |
| "Why is our pipeline slow / failing?" | cicd-architecture-auditor |
| Concern | Architecture | Widget Perf | State Mgmt | Runtime Perf | Atomic Design | Testability | CI/CD |
|---|---|---|---|---|---|---|---|
| Layer separation | ✅ | ||||||
| Modularization | ✅ | ||||||
| Widget rebuild cost | ✅ | ||||||
| Scroll / layout performance | ✅ | ||||||
| BLoC/Cubit/Riverpod patterns | ✅ | ||||||
| State rebuild propagation | ✅ | ||||||
| Isolate / async efficiency | ✅ | ||||||
| Caching / memoization | ✅ | ||||||
| Atomic hierarchy | ✅ | ||||||
| Design token adoption | ✅ | ||||||
| DI / constructor injection | ✅ | ||||||
| Bloc test readiness | ✅ | ||||||
| Widget / golden test readiness | ✅ | ||||||
| Pipeline job topology | ✅ | ||||||
| Test sharding / caching | ✅ | ||||||
| Platform build signing | ✅ | ||||||
| Release automation | ✅ |
Build a new feature:
"Build a product listing screen using clean architecture with BLoC state management and Firebase Firestore. Include error handling and pagination." → use
flutter-scalable-app
Audit architecture:
"Review the architecture of this Flutter app. Detect layer violations, inconsistent state management, and scalability risks." → use
architecture-auditor
Diagnose UI jank:
"The product list screen drops frames during scroll. Analyze widget rebuild patterns and layout costs." → use
widget-performance-analyzer
Audit state management:
"We use BLoC in some features and Riverpod in others. Audit the state management architecture and recommend a unification strategy." → use
state-management-auditor
Diagnose slow data operations:
"Our app freezes when processing large datasets. Audit isolate usage and async efficiency." → use
runtime-performance-auditor
Audit design system:
"We have duplicate button and text components across features. Audit our atomic design compliance and token usage." → use
atomic-design-system-auditor
Audit testability:
"We can't write unit tests for our Blocs and widget tests always need real repositories. Audit what's blocking us." → use
testability-architecture-auditor
Audit CI/CD pipeline:
"Our pipeline takes 40 minutes and fails intermittently. Audit the job topology, caching, and test sharding configuration." → use
cicd-architecture-auditor
flutter/
├── README.md
├── flutter-scalable-app/
│ ├── SKILL.md
│ └── references/
│ ├── architecture.md
│ ├── bloc-patterns.md
│ ├── dependency-injection.md
│ ├── firebase.md
│ ├── navigation.md
│ ├── performance.md
│ └── stitch-to-flutter.md
├── architecture-auditor/
│ ├── SKILL.md
│ └── references/
│ ├── architecture.md
│ ├── bloc-patterns.md
│ ├── dependency-injection.md
│ └── navigation.md
├── widget-performance-analyzer/
│ ├── SKILL.md
│ └── references/
│ ├── rebuild-patterns.md
│ ├── savelayer-and-compositing.md
│ ├── scroll-performance.md
│ └── layout-cost.md
├── state-management-auditor/
│ ├── SKILL.md
│ └── references/
│ ├── state-patterns.md
│ ├── scoping-and-lifecycle.md
│ ├── rebuild-efficiency.md
│ └── isolation-and-coupling.md
├── runtime-performance-auditor/
│ ├── SKILL.md
│ └── references/
│ ├── isolate-patterns.md
│ ├── async-efficiency.md
│ ├── memoization-and-caching.md
│ └── event-loop-health.md
├── atomic-design-system-auditor/
│ ├── SKILL.md
│ └── references/
│ ├── atomic-hierarchy.md
│ ├── token-usage.md
│ ├── component-composition.md
│ └── duplication-detection.md
├── testability-architecture-auditor/
│ ├── SKILL.md
│ └── references/
│ ├── unit-test-architecture.md
│ ├── bloc-test-patterns.md
│ ├── widget-test-patterns.md
│ └── golden-test-prerequisites.md
└── cicd-architecture-auditor/
├── SKILL.md
└── references/
├── diagnosis.md
├── testing.md
├── builds.md
├── quality.md
├── release.md
└── gh-actions.md