MCP Hub
Back to servers

Mobile Development Validator

A robust guardian for mobile development that enforces structured task planning, architectural integrity, and UI/UX best practices across Flutter, React Native, iOS, and Android platforms.

Stars
1
Tools
14
Updated
Dec 22, 2025

Mobile MCP Validator

🛡️ Mobile MCP Validator

Stop AI "Vibe Coding" — Enforce Disciplined Mobile Development

Quick StartFeaturesPlatformsToolsUI/UXAuthor

License Dart Platforms Tests


🎯 The Problem I Solved

After 8 years of mobile development, I noticed a recurring issue with AI coding assistants:

"Vibe Coding" — AI agents that write scattered, unplanned code that breaks things, ignores patterns, and produces amateur-looking UIs.

Sound familiar? The AI:

  • ❌ Jumps into coding without a plan
  • ❌ Changes things that shouldn't be changed
  • ❌ Ignores your architecture patterns
  • ❌ Creates inconsistent, unprofessional UIs
  • ❌ Forgets to handle loading, error, and empty states
  • ❌ Produces files that are 500+ lines long

I built Mobile MCP Validator to fix this.


💡 The Solution

Mobile MCP Validator is an MCP (Model Context Protocol) server for Cursor IDE that acts as a guardian for your mobile codebase.

It forces AI agents to:

BeforeAfter (with MCP)
Jump straight into coding✅ Create a task plan first
Make random changes✅ Validate every change
Ignore file size limits✅ Block files over 300 lines
Forget UI states✅ Require loading/error/empty/success
Break existing code✅ Detect breaking changes
Produce ugly UIs✅ Follow UI/UX best practices

📱 Supported Platforms

Works with ALL mobile development frameworks:

PlatformLanguagesLinter Integration
💙 FlutterDartflutter analyze
⚛️ React NativeJavaScript, TypeScriptESLint
🍎 iOS NativeSwift, Objective-CSwiftLint
🤖 Android NativeKotlin, Javaktlint

🚀 Quick Start

1. Clone the Repository

git clone https://github.com/emran1993/mobile-mcp-validator.git
cd mobile-mcp-validator
dart pub get

2. Add to Cursor

Add to your ~/.cursor/mcp.json:

{
  "mcpServers": {
    "mobile-mcp-validator": {
      "command": "dart",
      "args": ["run", "/path/to/mobile-mcp-validator/bin/mobile_mcp_validator.dart"]
    }
  }
}

3. Restart Cursor

That's it! The MCP is now protecting your codebase.


✨ Features

🎯 Task Planning (No More Random Changes)

The AI must create a plan before coding:

User: "Add a settings screen"

AI calls: create_task_plan()

→ Task 1: Analyze existing patterns
→ Task 2: Create settings controller  
→ Task 3: Build settings screen with all UI states
→ Task 4: Run validation
→ Task 5: Verify code quality

🚫 Blocking Errors (Forces Fixes)

When the AI makes a mistake, it gets blocked:

{
  "success": false,
  "isBlocking": true,
  "message": "BLOCKING: File exceeds 300 lines (found: 425)",
  "fixSuggestion": "Split into smaller modules..."
}

The AI cannot proceed until it fixes the issue.

🔍 Breaking Change Detection

Before modifying existing code:

  1. AI takes a "snapshot" of the current state
  2. After changes, compares to detect unintended modifications
  3. Alerts if classes, methods, or APIs were accidentally removed

🎨 UI/UX Best Practices

Comprehensive guidelines built-in:

  • Design Principles — Visual hierarchy, spacing, consistency
  • Color Systems — Palettes, dark mode, contrast
  • Typography — Font scales, weights, readability
  • 8pt Grid System — Consistent spacing
  • Animations — Timing, easing, micro-interactions
  • Accessibility — Touch targets, screen readers, contrast

🛠️ Tools

Task Planning

ToolWhat It Does
create_task_planCreates checklist before coding
validate_task_completionEnsures all tasks are done
update_task_statusTracks progress

Code Quality

ToolWhat It Does
validate_file_sizeBlocks files over 300 lines
validate_function_complexityBlocks functions over 50 lines
validate_widget_complexityBlocks build methods over 120 lines
find_business_logic_in_widgetsDetects logic leaking into UI

Breaking Changes

ToolWhat It Does
snapshot_ui_stateCaptures state before changes
detect_breaking_changesFinds unintended modifications
find_affected_usagesLists all code using a file

Platform Analysis

ToolWhat It Does
detect_project_frameworkAuto-detects Flutter/RN/iOS/Android
run_project_analysisRuns correct linter automatically
validate_ios_configChecks Info.plist, Podfile
validate_android_configChecks AndroidManifest, Gradle

🎨 UI/UX Guidelines

The MCP includes comprehensive resources for building beautiful apps:

ResourceDescription
uiux://design-principlesCore principles for professional UI
uiux://color-systemColors, themes, dark mode
uiux://typographyFont selection and hierarchy
uiux://spacing-layout8pt grid system
uiux://animationsMotion and micro-interactions
uiux://accessibilityInclusive design guidelines
uiux://platform-guidelinesiOS HIG & Material Design
uiux://common-mistakesAnti-patterns to avoid

⚙️ Configuration

Create flutter_guardian.yaml in your project:

rules:
  max_file_lines: 300
  max_build_method_lines: 120
  max_function_lines: 50

architecture:
  pattern: feature_first
  require_separation: true

testing:
  require_ui_states:
    - loading
    - error
    - empty
    - success

🔒 Pre-commit Hook

Validate code before every commit:

dart run mobile_mcp_validator:pre_commit --install

👨‍💻 About the Author

Emran Shiekh Ahmad

Senior Software Engineer @ SolarEdge

🚀 8+ years of mobile development experience

📱 Specializing in Flutter, iOS, and Android

💡 Passionate about code quality, architecture, and developer experience

🎯 Built this MCP to solve the "vibe coding" problem I kept seeing with AI assistants

Why I Built This

"After years of building mobile apps, I've learned that the difference between amateur and professional code isn't just functionality — it's discipline. Planning before coding. Consistent patterns. Clean architecture. Beautiful, accessible UIs.

When AI assistants started helping with code, I noticed they often lacked this discipline. They'd jump into coding without a plan, create massive files, ignore patterns, and produce inconsistent UIs.

Mobile MCP Validator is my solution. It brings the discipline of a senior developer to AI-assisted coding. It's the guardian I wish I had when I started — and now it's available for everyone."

Emran Shiekh Ahmad


🌟 Star This Project

If you find this useful, please ⭐ star this repository — it helps others discover it!


📄 License

MIT License — see LICENSE for details.

Copyright © 2025 Emran Shiekh Ahmad


🤝 Contributing

Contributions are welcome! Please:

  1. Fork the repository
  2. Create a feature branch
  3. Make your changes
  4. Run tests: dart test
  5. Submit a pull request

Built with ❤️ by Emran Shiekh Ahmad
Senior Software Engineer @ SolarEdge | 8+ Years Mobile Development

GitHubLinkedIn

Reviews

No reviews yet

Sign in to write a review