Skip to content
MCP server

Flutter MCP Toolkit

Let an AI assistant inspect, tap through and hot reload a running Flutter debug app

Runs locallyStatus and history
Install
Pricing
Open source
Platforms
CLI, macOS
Runs
Locally
Version
7.1.1
Listing updated
2 days ago

About Flutter MCP Toolkit

Flutter MCP Toolkit is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.

This server connects a Flutter app running in debug mode to an assistant so it can check changes against runtime evidence instead of guessing. After adding the mcp_toolkit package and bootstrapping it in the app, the assistant can capture screenshots, read app errors, dump layer, semantics, render and focus trees, hot reload or restart, and evaluate Dart expressions.

Interaction follows a Playwright-style model: take a semantic snapshot, then pass a ref such as s_N to tools that tap, long press, enter text, scroll, swipe or drag. A reveal_search helper scrolls to find off-screen targets, and a fused hot-reload-and-capture command returns a screenshot, fresh snapshot and errors in one call. Apps can register their own tools and resources at runtime through a dynamic registry.

Tool names in v3 carry an fmt_ prefix. The same core backs a CLI (flutter-mcp-toolkit, alias fmtk) with validate-runtime, doctor, snapshot and bundle commands, and an explicit target id avoids mix-ups when several debug apps run. The server ships as an OCI image over stdio.

Install

Run in a terminal

claude mcp add flutter-mcp-toolkit -- docker run -i --rm ghcr.io/arenukvern/flutter-mcp-toolkit:7.1.1

Add to claude_desktop_config.json

{
  "mcpServers": {
    "flutter-mcp-toolkit": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/arenukvern/flutter-mcp-toolkit:7.1.1"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "flutter-mcp-toolkit": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/arenukvern/flutter-mcp-toolkit:7.1.1"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "flutter-mcp-toolkit": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/arenukvern/flutter-mcp-toolkit:7.1.1"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "flutter-mcp-toolkit": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/arenukvern/flutter-mcp-toolkit:7.1.1"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "flutter-mcp-toolkit": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "ghcr.io/arenukvern/flutter-mcp-toolkit:7.1.1"
      ]
    }
  }
}

Tools this server provides (5)

Taken from the project documentation; the running server may differ.

fmt_tap_widget
Tap a widget by snapshot ref.
fmt_hot_reload_and_capture
Hot reload and return a screenshot, fresh snapshot and errors.
fmt_list_client_tools_and_resources
List tools and resources the app registered at runtime.
fmt_client_tool
Call an app-registered tool.
fmt_client_resource
Read an app-registered resource.

What it does

Semantic snapshots
Captures the UI tree so interaction tools can target elements by ref.
UI interaction
Tap, long press, type, scroll, swipe and drag on the running app.
Hot reload with capture
Reloads and returns a screenshot, snapshot and errors together.
Runtime diagnostics
Reads app errors and dumps layer, semantics, render and focus trees.
Dynamic app tools
Your app registers custom MCP tools and resources at runtime.
Multi-target safety
An explicit target id selects the app when several debug apps run.

Good for

  • Verify a UI fix on a live Flutter app
  • Reproduce a bug by driving the screen from an assistant
  • Read runtime errors after a hot reload
  • Expose app-specific actions as tools for an agent
  • Run scripted E2E checks from the CLI

Pricing

Open source

Open-source repository; no price listed.

Prices change. Check docs.page before you buy.

Flutter MCP Toolkit FAQ

Does it work with release builds?

The docs describe debug apps; it connects to a running Flutter debug app.

What is the tool prefix?

In v3 MCP tool names start with fmt_, for example fmt_tap_widget.

Is there a command line?

Yes, flutter-mcp-toolkit with the short alias fmtk, including validate-runtime and doctor.

How do I install it?

Add mcp_toolkit to the app and run the server from its OCI image over stdio.

Is Flutter MCP Toolkit free?

Yes. Flutter MCP Toolkit is open source, so you can use and self-host it for free.

How do I add Flutter MCP Toolkit to Claude Code?

Run this command in your terminal: claude mcp add flutter-mcp-toolkit -- docker run -i --rm ghcr.io/arenukvern/flutter-mcp-toolkit:7.1.1

What are the best alternatives to Flutter MCP Toolkit?

The closest alternatives to Flutter MCP Toolkit are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Flutter MCP Toolkit

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools