Why This Repo Matters
Originally this project was about creating Dart MCP server to debug Flutter apps (when there was no official library existed).
When official Dart MCP landed - it shifted my priority and implementation principles.
At first, I was thinking to abandon project, however while I was developing next version, I came to unexpected idea - to redesign tools and resources as Toolkit - to allow agents to develop it for any app, and share it from the app, not ship with server.
That lead me to continue working on this project from that point - and that's the main difference between official Dart Server and this MCP Toolkit - they supposed to work side by side.
Most of the changes / motivation is described in Changelog for v2
And the current goal: to make Flutter debugging and automation usable and customizable for AI Agents for all Flutter applicaitons.
The Problem
Generic MCP or Dart-only tooling can expose low-level VM data, but real Flutter debugging usually needs:
- app-level errors with useful summaries
- UI screenshots and view metadata
- fast actions like hot reload/hot restart
- app-specific tools that know your domain
What This Repo Adds
- A Dart MCP server (
flutter-mcp-toolkit-server) focused on Flutter workflows. - A CLI (
flutter-mcp-toolkit) for deterministic automation, snapshots, and CI. flutter_mcp_toolkitpackage for in-app extensions.- Dynamic tool/resource registration from the running app.
- Target-resolution and retry guidance for multi-app debug sessions.
Get started in 4 steps
# 1. Install the binary
curl -fsSL https://raw.githubusercontent.com/Arenukvern/flutter-mcp-toolkit/main/install.sh | bash
# 2. Add the toolkit to your Flutter app
cd my-flutter-app
flutter-mcp-toolkit codegen-init # adds flutter_mcp_toolkit + emits main.dart snippet
# 3. Install skills for your AI agent
flutter-mcp-toolkit init claude-code # or: cursor | codex | cline | all
# 4. Run the app
flutter run --debug
That's it. Your AI agent can now inspect and drive the running app.