Menü
Use when you have a written implementation plan to execute with review checkpoints
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Basierend auf der SOC-Berufsklassifikation
Creates a new Flutter project with Clean Architecture, domain pattern boilerplate, and production-ready setup
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation using Clean Architecture principles.
Use when facing 2+ independent Flutter tasks that can be worked on without shared state or sequential dependencies
Use when you have a design or requirements for a multi-step Flutter feature, before touching code
Use when executing Flutter implementation plans with independent tasks in the current session
Use when writing tests for Flutter code - follows priority-based testing (Repository → State → Widget) after implementation
| name | flutter-executing |
| description | Use when you have a written implementation plan to execute with review checkpoints |
Load plan, review critically, execute tasks in batches, run Flutter verifications, report for review between batches.
Core principle: Batch execution with checkpoints for review.
Announce at start: "I'm using the flutter-executing skill to implement this plan."
Read plan file
cat docs/plans/YYYY-MM-DD-<feature>-plan.md
Check dependencies first
cat pubspec.yaml
Install new dependencies if needed
flutter pub get
Review critically:
If concerns: Raise them before starting
If no concerns: Create todo entries (task tracking tool) and proceed
Default: First 3 tasks
For each task:
flutter analyze lib/features/<feature>/
git add <files>
git commit -m "<conventional commit message>"
After completing 3 tasks, run full verification:
# Static analysis
flutter analyze
# Expected: No issues found!
# Run tests (if test files exist)
flutter test test/features/<feature>/
# Expected: All tests passed!
# Quick build check (optional — full build runs once in flutter-finishing)
flutter build <target> --debug # apk / ios / web / windows … per project platform
# Expected: Built successfully
When batch complete, report:
## Batch N Complete
**Tasks completed:**
1. ✅ Task 1: [description]
2. ✅ Task 2: [description]
3. ✅ Task 3: [description]
**Verification:**
- flutter analyze: No issues found
- flutter test: X/X passed
**Files created/modified:**
- lib/features/.../entity.dart
- lib/features/.../repository.dart
- ...
**Ready for feedback.**
Based on feedback:
After all tasks complete and verified:
Run final verification:
flutter analyze
flutter test
flutter build <target> --debug # per project platform (apk/ios/web/…)
Announce: "I'm using the flutter-finishing skill to complete this work."
REQUIRED SUB-SKILL: Use flutter-craft:flutter-finishing
| Stage | Command | Expected Output |
|---|---|---|
| Per-file | flutter analyze lib/path/file.dart | No issues found |
| Per-batch | flutter analyze lib/features/<feature>/ | No issues found |
| Test | flutter test test/features/<feature>/ | All tests passed |
| Final | flutter build <target> --debug | Built successfully |
STOP executing immediately when:
flutter analyze shows errors (not just warnings)Ask for clarification rather than guessing.
Before executing, verify plan follows this order:
1. Domain Layer tasks first
├── Entities
├── Repository interfaces
└── UseCases
2. Data Layer tasks second
├── Models
├── DataSources
└── Repository implementations
3. Presentation Layer tasks third
├── State Management
├── Widgets
└── Screens
4. Test tasks after implementation
├── Repository tests (priority 1)
├── State tests (priority 2)
└── Widget tests (priority 3)
If plan doesn't follow this order, raise concern before starting.
After EACH batch, you MUST invoke: → flutter-craft:flutter-verification (verify before continuing)
You MUST invoke: → flutter-craft:flutter-finishing (present completion options)