Flutter MCP Service

Flutter MCP Service v2.0 enables intelligent Flutter tooling through the Model Context Protocol (MCP). It provides fast access to Flutter/Dart docs, code analysis, performance insights, and project-wide guidance, all while managing tokens and caching for efficient, repeatable interactions with your development workbench.

How to use

You interact with the Flutter MCP Service through an MCP client. Start tasks that analyze widgets, search documentation, validate code against official Flutter docs, and generate tests or architecture guidance. Each tool is designed to help you improve code quality, reduce debugging time, and align your Flutter projects with best practices. You can combine multiple tools in a workflow, such as analyzing a widget tree for performance, then requesting improvements and generating tests for the optimized code.

How to install

Prerequisites: ensure you have Node.js and npm installed on your system. You may also need Python or other runtime environments depending on your local setup, but the core service runs through Node.js.

# Clone the service repository
git clone https://github.com/dvillegastech/flutter_mcp_2.git
cd flutter_mcp_service

# Install dependencies
npm install

# Initialize cache directory (optional but recommended)
mkdir -p .cache

# Run a health check to verify the service readiness
npm run health-check

# Start the MCP service
npm start

Configuration and integration

To expose the Flutter MCP Service to an MCP client, add a server entry in your MCP configuration. The following example shows a local stdio (child process) setup that launches the service from its index script.

{
  "mcpServers": {
    "flutter-mcp": {
      "command": "node",
      "args": ["/absolute/path/to/flutter_mcp_service/src/index.js"]
    }
  }
}

Troubleshooting and notes

If you encounter startup or health issues, verify that dependencies are installed, the cache directory exists, and the start command is executed from the project root. Ensure the path to the index script is correct and that the environment where the service runs has access to necessary runtime resources.

Examples of workflows

Common workflows involve analyzing a widget, then performing a series of actions to improve performance and reliability. For example, you can analyze a widget tree for performance, request improvement suggestions, and then generate tests for the optimized code to ensure regressions are caught early.

Notes on usage patterns

You can run health checks, observe cache statistics, and use the recommended tools in sequence to build confidence in your Flutter projects. The service is designed to support multi-tool workflows, enabling you to inspect docs, analyze code, and validate architecture in a cohesive manner.

Available tools

flutter_status

Check service health and cache statistics.

flutter_search

Universal search across Flutter/Dart documentation, packages, and examples.

flutter_analyze

Smart Flutter documentation fetcher and code analyzer.

analyze_widget

Analyze Flutter widget code for best practices, performance, and accessibility.

validate_flutter_docs

Validate code against official Flutter documentation.

analyze_pub_package

Analyze packages from pub.dev for quality and compatibility.

suggest_improvements

Get improvement suggestions based on Flutter best practices.

analyze_performance

Analyze Flutter widget tree for performance issues.

analyze_architecture

Analyze project architecture compliance.

analyze_bundle_size

Analyze app bundle size with optimization recommendations.

generate_tests

Generate comprehensive Flutter tests.

trace_state

Trace state flow and rebuilds in Flutter widgets.

generate_clean_architecture

Generate Clean Architecture structure.

generate_l10n

Generate localization setup with ARB files.

monitor_performance

Generate performance monitoring setup.

diagnose_render_issues

Diagnose and fix rendering issues.

analyze_test_coverage

Analyze test coverage with recommendations.