vp-k/flutter-craft/flutter-executing

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

Use when you have a written implementation plan to execute with review checkpoints

Install

npx vskill@latest install vp-k/flutter-craft/flutter-executing

SKILL.md

Executing Flutter Implementation Plans

Overview

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."

The Process

Step 1: Load and Review Plan
  1. Read plan file

``bash cat docs/plans/YYYY-MM-DD-<feature>-plan.md ``

  1. Check dependencies first

``bash cat pubspec.yaml ``

  1. Install new dependencies if needed

``bash flutter pub get ``

  1. Review critically:
  • Are file paths correct?
  • Is layer order followed (Domain → Data → Presentation)?
  • Are there any gaps in the plan?
  1. If concerns: Raise them before starting
  2. If no concerns: Create todo entries (task tracking tool) and proceed
Step 2: Execute Batch

Default: First 3 tasks

For each task:

  1. Mark as in_progress in the task tracker
  2. Create/modify files as specified
  3. Write complete code (from plan)
  4. Run verification:

``bash flutter analyze lib/features/<feature>/ ``

  1. Commit:

``bash git add <files> git commit -m "<conventional commit message>" ``

  1. Mark as completed in the task tracker
Step 3: Verify Batch

After completing 3 tasks, run full verification:

bash
# 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
Step 4: Report

When batch complete, report:

markdown
## 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.**
Step 5: Continue

Based on feedback:

  • Apply changes if needed
  • Execute next batch of 3 tasks
  • Repeat until all tasks complete
Step 6: Complete Development

After all tasks complete and verified:

  1. Run final verification:

``bash flutter analyze flutter test flutter build <target> --debug # per project platform (apk/ios/web/…) ``

  1. Announce: "I'm using the flutter-finishing skill to complete this work."
  1. REQUIRED SUB-SKILL: Use flutter-craft:flutter-finishing
  • Verify tests
  • Present 4 options (merge/PR/keep/discard)
  • Execute choice
  • Cleanup worktree if applicable

Flutter-Specific Verification Commands

StageCommandExpected Output
Per-fileflutter analyze lib/path/file.dartNo issues found
Per-batchflutter analyze lib/features/<feature>/No issues found
Testflutter test test/features/<feature>/All tests passed
Finalflutter build <target> --debugBuilt successfully

When to Stop and Ask for Help

STOP executing immediately when:

  • flutter analyze shows errors (not just warnings)
  • Tests fail with unclear cause
  • Missing dependency not in pubspec.yaml
  • Plan has unclear instructions
  • File path in plan doesn't match project structure
  • State management pattern unclear

Ask for clarification rather than guessing.

Layer Order Verification

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, optional)
└── Golden tests (priority 4, optional, visual regression)

If plan doesn't follow this order, raise concern before starting.

REQUIRED SUB-SKILL

During Execution

After EACH batch, you MUST invoke: → flutter-craft:flutter-verification (verify before continuing)

After All Tasks Complete

You MUST invoke: → flutter-craft:flutter-finishing (present completion options)

Remember

  • Review plan critically first - Don't blindly execute
  • Follow layer order - Domain → Data → Presentation
  • Run flutter analyze per batch - Catch errors early
  • Commit after each task - Atomic commits
  • Report and wait - Don't continue without feedback
  • Stop when blocked - Don't guess, ask

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
14
Repository forks
6
7-day trend score
10

Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 7 May 2026, 5 months ago.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/vp-k/flutter-craft/flutter-executing/badge)](https://verified-skill.com/skills/vp-k/flutter-craft/flutter-executing)

Something wrong with this skill or its page? Tell us.