Flutter AI Skills Template

A ready-to-use Flutter project template with AI skills pre-installed for both Claude and Cursor.

The project was initialized with flutter create ., and the skills below are already wired up for both tools — no bootstrap step required.

This template standardizes:

• Clean Architecture
• Feature-first development
• API integrations
• Authentication
• Local storage
• Caching
• Testing
• Pull Request Reviews

---

Repository Structure

flutter_ai_skills/
├── README.md
├── AGENTS.md
├── CLAUDE.md
│
├── .claude/
│   └── skills/
│       ├── project-bootstrap/
│       ├── architecture/
│       ├── api-standards/
│       ├── auth-standards/
│       ├── data-storage/
│       ├── caching/
│       ├── unit-testing/
│       └── flutter-pr-review/
│
├── .cursor/
│   └── skills/
│       └── ... (mirrors .claude/skills)
│
├── lib/
│   └── main.dart
├── test/
├── android/ ios/ web/ linux/ macos/ windows/
└── pubspec.yaml

---

Prerequisites

Install either:

Flutter

flutter --version

FVM

fvm --version

---

Getting Started

This repo is a Flutter project, not just a skills library. Use it as a template for new apps:

git clone <this-repo-url> banking_app
cd banking_app
flutter pub get

Cloning only renames the folder — the project is still identified as flutter_ai_skills everywhere else (pubspec.yaml name, Android applicationId, iOS bundle identifier, desktop app names, etc.). To actually rename the app, use the rename package:

dart pub global activate rename
rename setAppName --value "Banking App"
rename setBundleId --value "com.example.banking_app"

Or update these manually if you prefer not to install the tool:

• pubspec.yaml → name:
• android/app/build.gradle → applicationId
• ios/Runner.xcodeproj / Info.plist → bundle identifier
• linux/, macos/, windows/ → executable/display name

Open with Cursor

cursor .

Open with Claude

claude

Both tools pick up their respective skills automatically — Cursor from .cursor/skills/, Claude from .claude/skills/ (driven by CLAUDE.md at the project root).

---

Available Skills

Skill	Purpose
project-bootstrap	Project initialization
architecture	Architecture and feature generation
api-standards	APIs, repositories, models, datasources
auth-standards	Authentication and authorization
data-storage	Hive and Secure Storage
caching	Offline support and caching
unit-testing	Unit, widget and integration testing
flutter-pr-review	Pull request review

---

Developer Workflow

Step 1

Initialize the project:

Initialize this project

The AI automatically:

• Loads project-bootstrap
• Loads architecture
• Creates the project foundation

---

Step 2

Create a feature:

Create Customer feature

The AI automatically:

• Creates domain layer
• Creates data layer
• Creates presentation layer
• Creates providers
• Creates routing
• Creates tests when required

---

Step 3

Create authentication:

Create Login feature

The AI automatically applies:

• architecture
• auth-standards
• data-storage
• unit-testing

---

Step 4

Add caching:

Add caching to Customer feature

The AI automatically applies:

• caching

---

Step 5

Generate tests:

Generate tests for Customer feature

The AI automatically applies:

• unit-testing

---

Step 6

Review code:

Review current branch changes

The AI automatically applies:

• flutter-pr-review

---

Example Prompts

Initialize this project

Create Login feature

Create Customer feature

Create Dashboard feature

Add caching to Customer feature

Generate tests for Customer feature

Review current branch changes

---

Architecture

The architecture specification is defined in:

.claude/skills/architecture/SKILL.md
.cursor/skills/architecture/SKILL.md

This is the source of truth for:

• Clean Architecture
• Feature First Structure
• Dependency Rules
• Riverpod
• GoRouter
• Dio
• Hive
• Secure Storage
• Testing Standards

All generated code must follow this specification.

---

Project Priorities

1. Security
2. Scalability
3. Clean Architecture
4. Testability
5. Maintainability

All skills and generated code must follow these priorities.