mcpbeat Sign in

Makepad 2 0 App Structure Skill for Claude

| AppMain, handle_event, handle_actions, ScriptVm, from_script_mod, makepad boilerplate, makepad new project, makepad cargo, Cargo.toml setup, hot reload, --hot, live reload, wasm deploy, cargo makepad, media plugin, audio_output, audio_input, AudioBuffer, cx.audio, makepad audio, 音频, 应用结构, 入门, 新项目, 脚手架, 启动, 热重载, 部署

5k tokens
context cost
the whole folder, loaded on every use
3
files
instructions only
0
copies elsewhere
how many repositories repackaged it
745
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/ZhangHanDong/makepad-skills --skill makepad-2.0-app-structure

What comes with it

5 790 bytes besides the instruction
references/app-boilerplate.md
references/rust-splash-integration.md

The instruction itself

41 sections, as written by the author

Makepad 2.0 App Structure Skill

> Version: makepad-widgets (dev branch) | Last Updated: 2026-03-03

Overview

A Makepad 2.0 app combines Rust code with Splash scripting. The Rust side handles app lifecycle, event routing, and business logic. The Splash side defines UI structure, templates, and inline interactions.

Documentation

Refer to the local files for detailed documentation:

  • ./references/app-boilerplate.md - Complete working app template with Cargo.toml
  • ./references/rust-splash-integration.md - Rust ↔ Splash communication patterns

IMPORTANT: Documentation Completeness Check

Before answering questions, Claude MUST:

  • Read the relevant reference file(s) listed above
  • If file read fails, answer based on SKILL.md patterns + built-in knowledge

Minimal App Template

Cargo.toml

[package]
name = "my-app"
version = "0.1.0"
edition = "2024"

[dependencies]
makepad-widgets = { path = "../path/to/makepad/widgets" }

src/app.rs (or src/main.rs)

use makepad_widgets::*;

app_main!(App);

script_mod! {
    use mod.prelude.widgets.*

    let state = {
        counter: 0
    }
    mod.state = state

    startup() do #(App::script_component(vm)){
        ui: Root{
            on_startup: ||{
                ui.main_view.render()
            }
            main_window := Window{
                window.inner_size: vec2(420, 300)
                body +: {
                    main_view := View{
                        width: Fill height: Fill
                        flow: Down spacing: 12
                        align: Center
                        on_render: ||{
                            Label{
                                text: "Count: " + state.counter
                                draw_text.text_style.font_size: 24
                            }
                        }
                    }
                    increment_button := Button{
                        text: "Increment"
                    }
                }
            }
        }
    }
}

impl App {
    fn run(vm: &mut ScriptVm) -> Self {
        crate::makepad_widgets::script_mod(vm);
        App::from_script_mod(vm, self::script_mod)
    }
}

#[derive(Script, ScriptHook)]
pub struct App {
    #[live]
    ui: WidgetRef,
}

impl MatchEvent for App {
    fn handle_actions(&mut self, cx: &mut Cx, actions: &Actions) {
        if self.ui.button(cx, ids!(increment_button)).clicked(actions) {
            script_eval!(cx, {
                mod.state.counter += 1
                ui.main_view.render()
            });
        }
    }
}

impl AppMain for App {
    fn handle_event(&mut self, cx: &mut Cx, event: &Event) {
        self.match_event(cx, event);
        self.ui.handle_event(cx, event, &mut Scope::empty());
    }
}

Key Components Explained

1. app_main!(App)

Registers App as the application entry point. MUST be called at module level.

2. script_mod! { ... }

Defines the Splash script that runs at startup. Contains:

  • use mod.prelude.widgets.* - Import widget definitions
  • State definitions (let state = {...})
  • Functions and templates
  • startup() do #(App::script_component(vm)){...} - UI construction

3. App::run(vm) - Initialization Order

CRITICAL: Registration order matters!

fn run(vm: &mut ScriptVm) -> Self {
    // 1. Register theme (optional, for light/dark theme)
    crate::makepad_widgets::theme_mod(vm);
    script_eval!(vm, { mod.theme = mod.themes.light });

    // 2. Register base widgets
    crate::makepad_widgets::widgets_mod(vm);

    // 3. Register custom widget modules (if any)
    // crate::my_widgets::script_mod(vm);

    // 4. Build app from script module
    App::from_script_mod(vm, self::script_mod)
}

Simplified (without theme selection):

fn run(vm: &mut ScriptVm) -> Self {
    crate::makepad_widgets::script_mod(vm);   // Registers both theme + widgets
    App::from_script_mod(vm, self::script_mod)
}

4. #[derive(Script, ScriptHook)] on App struct

#[derive(Script, ScriptHook)]
pub struct App {
    #[live]
    ui: WidgetRef,    // The widget tree root
}
  • Script - Enables Splash integration (replaces old Live)
  • ScriptHook - Enables lifecycle hooks (replaces old LiveHook)
  • #[live] - Field settable from DSL

5. MatchEvent for Action Handling

impl MatchEvent for App {
    fn handle_actions(&mut self, cx: &mut Cx, actions: &Actions) {
        // Check button clicks
        if self.ui.button(cx, ids!(my_button)).clicked(actions) {
            // Handle click
        }
        // Check text input changes
        if let Some(text) = self.ui.text_input(cx, ids!(my_input)).changed(actions) {
            log!("Input: {}", text);
        }
    }
}

6. AppMain for Event Dispatch

impl AppMain for App {
    fn handle_event(&mut self, cx: &mut Cx, event: &Event) {
        self.match_event(cx, event);  // MUST call for MatchEvent to work
        self.ui.handle_event(cx, event, &mut Scope::empty());
    }
}

Rust → Splash Communication

script_eval! - Update state and trigger renders

script_eval!(cx, {
    mod.state.counter += 1
    ui.main_view.render()
});

script_apply_eval! - Patch widget properties

script_apply_eval!(cx, self.ui, {
    title.text: "New Title"
    subtitle.draw_text.color: #f00
});

Splash → Rust Communication

Inline Event Handlers in script_mod!

Button{
    text: "Click"
    on_click: || {
        // Splash code runs here
        state.count += 1
        ui.display.render()
    }
}

TextInput{
    on_return: || {
        let text = ui.my_input.text()
        add_item(text)
        ui.my_input.set_text("")
    }
}

// Render callback
on_render: || {
    for i, item in items {
        ItemTemplate{label.text: item.name}
    }
}

Widget Access from Rust

// Buttons
self.ui.button(cx, ids!(button_name)).clicked(actions)

// Labels
self.ui.label(cx, ids!(label_name)).set_text(cx, "text")

// Text inputs
self.ui.text_input(cx, ids!(input_name)).text()

// Nested access
self.ui.label(cx, ids!(container.inner.title))

Running the App

# Development
cargo run -p my-app

# Development with hot reload (Splash changes apply without recompilation)
cargo run -p my-app -- --hot

# Release
cargo run -p my-app --release

# With cargo-makepad for mobile/web
cargo makepad run -p my-app

Command-Line Flags

| Flag | Description |

|------|-------------|

| --hot | Enable hot reload: watches script_mod! source files and auto-refreshes UI on save. Only affects Splash DSL; Rust code changes still need recompilation. |

| --stdin-loop | Studio mode: communicates with Makepad Studio via stdin/websocket. Used internally by Studio, not for manual use. |

| --linux-backend=<x11\|wayland> | (Linux only) Select windowing backend. |


Makepad Studio: Debugging Splash

Running Splash Scripts in Studio

Studio looks for a makepad.splash file in the project root. It can run Splash scripts directly via the Run List panel (uses start_script_run internally), without needing cargo run.

Debug Output

Use std.println() / std.print() inside Splash scripts — output appears in both Studio's Log View panel and terminal stdout:

// makepad.splash
std.println("debug: value = " + my_var);

Error Display

Script failures show in Studio's Log View with file path and error details:

script failed while evaluating makepad.splash: <error details>

Hub API (Studio-injected hub module)

When running under Studio, Splash scripts get a hub module:

| API | Description |

|-----|-------------|

| hub.run(env, cmd, args) | Launch subprocess from Splash |

| hub.set_run_items(items) | Register runnable items in Studio's Run List |

| hub.studio_ip | Studio's WebSocket address |

Current Limitations

  • No breakpoint debugging — Splash VM does not support breakpoints or stepping
  • No AST dump flag — Inspecting parse results requires adding logs in Rust source (script/)
  • Print-based debuggingstd.println() is the primary debugging tool

Audio API (Rust-only, not exposed to Splash)

Makepad 2.0 provides native audio I/O via the CxMediaApi trait. Audio callbacks run on a separate real-time thread.

Audio Output

use makepad_widgets::*;

impl AppMain for App {
    fn handle_event(&mut self, cx: &mut Cx, event: &Event) {
        match event {
            Event::Startup => {
                cx.audio_output(0, move |_info, buffer: &mut AudioBuffer| {
                    for frame in 0..buffer.frame_count() {
                        let sample = /* generate sample */;
                        buffer.set(0, frame, sample);
                        buffer.set(1, frame, sample);
                    }
                });
            }
            _ => {}
        }
        self.match_event(cx, event);
        self.ui.handle_event(cx, event, &mut Scope::empty());
    }
}

Audio Input

cx.audio_input(0, move |info: AudioInfo, buffer: &AudioBuffer| {
    let sample_rate = info.sample_rate;
    for frame in 0..buffer.frame_count() {
        let sample = buffer.get(0, frame);
    }
});

Key Audio Types (platform/src/audio.rs)

| Type | Description |

|------|-------------|

| AudioBuffer | Multichannel sample container (f32) |

| AudioInfo | Device ID, sample rate, timing |

| AudioDeviceDesc | Device metadata (name, type, channels) |

| AudioStreamSender/Receiver | Decoupled audio routing with adaptive buffering |

Cross-Thread Data Flow Pattern (Audio → UI)

Audio callbacks run on a real-time thread. Use atomics to pass data to the UI thread:

use std::sync::atomic::{AtomicU64, Ordering};
use std::sync::Arc;

let amplitude = Arc::new(AtomicU64::new(0));
let amp_clone = amplitude.clone();

cx.audio_output(0, move |_info, buffer: &mut AudioBuffer| {
    let mut sum = 0.0f32;
    for frame in 0..buffer.frame_count() {
        let s = buffer.get(0, frame);
        sum += s * s;
    }
    let rms = (sum / buffer.frame_count() as f32).sqrt();
    amp_clone.store(rms.to_bits() as u64, Ordering::Relaxed);
});

// UI thread: read on NextFrame
let rms = f32::from_bits(amplitude.load(Ordering::Relaxed) as u32);

IMPORTANT: Audio is NOT exposed to Splash scripting. All audio processing must happen in Rust.

Platform Support

| Platform | Audio Output | Audio Input | Implementation |

|----------|-------------|-------------|----------------|

| macOS/iOS | AudioUnit | AVCapture | apple_media.rs |

| Windows | WASAPI | WASAPI | windows_media.rs |

| Linux | ALSA/PulseAudio | ALSA | linux_media.rs |

| Web/WASM | WebAudio | MediaDevices | web_media.rs |

| Android | AAudio | NDK Camera2 | android_media.rs |

Reference: TeamTalk Example

examples/teamtalk/ demonstrates full P2P audio chat with cx.audio_input()/cx.audio_output(), resampling, and UDP streaming.


Best Practices

  • Registration order - Theme → Base widgets → Custom widgets → App module
  • Use script_eval! to bridge Rust actions to Splash state updates
  • Call self.match_event(cx, event) in handle_event (required for MatchEvent)
  • Use on_render for dynamic content, call .render() to trigger updates
  • Keep business logic in Rust, keep UI declarations in Splash
  • Use mod.state for app-wide state accessible from both Rust and Splash
  • Audio processing in Rust only - Use atomics for audio→UI data flow

macOS System Integration Patterns (learned 2026-03-31)

NSStatusBar (Menu Bar Icon)

Makepad has no built-in NSStatusBar API. Use makepad_objc_sys (same ObjC runtime as Makepad) for system tray integration:

use makepad_objc_sys::{msg_send, class, sel, sel_impl};
use makepad_objc_sys::runtime::{Object, Sel, YES, NO};

// Create status bar item
let status_bar: ObjcId = msg_send![class!(NSStatusBar), systemStatusBar];
let item: ObjcId = msg_send![status_bar, statusItemWithLength: -1.0f64];
let button: ObjcId = msg_send![item, button];
let () = msg_send![button, setTitle: str_to_nsstring("MIC")];

Critical gotchas:

  • Do NOT use objc2 crate — it creates a separate NSApplication instance that conflicts with Makepad's
  • show_in_dock(false) hides NSStatusItem — use LSUIElement=true in Info.plist instead
  • Menu target objects get autoreleased — use a global singleton target + sender.tag() for action dispatch
  • Do NOT call CGEventTapEnable(tap, true) — CGEvent taps are enabled by default on creation; calling Enable actually disables them

Audio Input — Device ID Required

cx.use_audio_inputs(&[]) means stop recording, not "use default device". Must get device ID from AudioDevicesEvent:

impl MatchEvent for App {
    fn handle_audio_devices(&mut self, _cx: &mut Cx, e: &AudioDevicesEvent) {
        // Save default device ID
        self.default_input = e.default_input().into_iter().next();
    }
}

// When starting recording:
if let Some(device_id) = self.default_input {
    cx.use_audio_inputs(&[device_id]);  // Start capture with specific device
}

// When stopping:
cx.use_audio_inputs(&[]);  // Empty = stop all audio input

CGEvent Tap — Global Hotkey

For global keyboard shortcuts (e.g., press-to-talk):

// CGEventGetFlags returns 0x80000 for Option key (not 0x20 as some docs suggest)
let key_mask: u64 = 0x080000; // kCGEventFlagMaskAlternate

// Use CFRunLoopCommonModes (not DefaultMode)
CFRunLoopAddSource(run_loop, source, kCFRunLoopCommonModes);

// Do NOT call CGEventTapEnable — it's already enabled
// CGEventTapEnable(tap, true);  // WRONG — this actually breaks it

// CFRunLoopRun blocks forever — put in dedicated thread
CFRunLoopRun();

Cross-Thread Communication Pattern

macos-sys (CFRunLoop thread) → crossbeam channel → Makepad timer poll (main thread)
Audio callback (RT thread)   → Arc<AtomicU64>    → NextFrame handler (main thread)
HTTP response               → MatchEvent handler → UI update via script_eval!

Timer poll pattern for receiving ObjC callbacks:

// In handle_timer (10ms interval):
let events: Vec<u64> = self.menu_rx.as_ref()
    .map(|rx| rx.try_iter().collect())
    .unwrap_or_default();
for action_id in events {
    self.handle_menu_action(cx, action_id);
}

Other skills for the same job

different authors, same section of the catalogue
Azure Kubernetes Automatic Readiness
by microsoft
vendor ×3

Assess Kubernetes workloads and cluster configuration for AKS Automatic compatibility. Identifies incompatibilities, generates fixes, and guides migration from AKS Standard to AKS Automatic. WHEN: migrate to AKS Automatic, check AKS Automatic readiness, validate manifests for Automatic, assess cluster for Automatic compatibility, fix deployment for Automatic compatibility, identify AKS Automatic migration blockers, is my cluster ready for AKS Automatic.

13k tokens
Capacity
by microsoft
vendor ×3

Discovers available Azure OpenAI model capacity across regions and projects. Analyzes quota limits, compares availability, and recommends optimal deployment locations based on capacity requirements. USE FOR: find capacity, check quota, where can I deploy, capacity discovery, best region for capacity, multi-project capacity search, quota analysis, model availability, region comparison, check TPM availability. DO NOT USE FOR: actual deployment (hand off to preset or customize after discovery), quota increase requests (direct user to Azure Portal), listing existing deployments.

6k tokens scripts
Customize
by microsoft
vendor ×3

Interactive guided deployment flow for Azure OpenAI models with full customization control. Step-by-step selection of model version, SKU (GlobalStandard/Standard/ProvisionedManaged), capacity, RAI policy (content filter), and advanced options (dynamic quota, priority processing, spillover). USE FOR: custom deployment, customize model deployment, choose version, select SKU, set capacity, configure content filter, RAI policy, deployment options, detailed deployment, advanced deployment, PTU deployment, provisioned throughput. DO NOT USE FOR: quick deployment to optimal region (use preset).

8k tokens
Deploy Model
by microsoft
vendor ×3

Unified Azure OpenAI model deployment skill with intelligent intent-based routing. Handles quick preset deployments, fully customized deployments (version/SKU/capacity/RAI policy), and capacity discovery across regions and projects. USE FOR: deploy model, deploy gpt, create deployment, model deployment, deploy openai model, set up model, provision model, find capacity, check model availability, where can I deploy, best region for model, capacity analysis. DO NOT USE FOR: listing existing deployments (use foundry_models_deployments_list MCP tool), deleting deployments, agent creation (use agent/create), project creation (use project/create).

26k tokens scripts
Preset
by microsoft
vendor ×3

Intelligently deploys Azure OpenAI models to optimal regions by analyzing capacity across all available regions. Automatically checks current region first and shows alternatives if needed. USE FOR: quick deployment, optimal region, best region, automatic region selection, fast setup, multi-region capacity check, high availability deployment, deploy to best location. DO NOT USE FOR: custom SKU selection (use customize), specific version selection (use customize), custom capacity configuration (use customize), PTU deployments (use customize).

9k tokens
Lamindb
by christophacham
×3

This skill should be used when working with LaminDB, an open-source data framework for biology that makes data queryable, traceable, reproducible, and FAIR. Use when managing biological datasets (scRNA-seq, spatial, flow cytometry, etc.), tracking computational workflows, curating and validating data with biological ontologies, building data lakehouses, or ensuring data lineage and reproducibility in biological research. Covers data management, annotation, ontologies (genes, cell types, diseases, tissues), schema validation, integrations with workflow managers (Nextflow, Snakemake) and MLOps platforms (W&B, MLflow), and deployment strategies.

22k tokens
Latchbio Integration
by christophacham
×3

Latch platform for bioinformatics workflows. Build pipelines with Latch SDK, @workflow/@task decorators, deploy serverless workflows, LatchFile/LatchDir, Nextflow/Snakemake integration.

12k tokens
Modal
by christophacham
×3

Run Python code in the cloud with serverless containers, GPUs, and autoscaling. Use when deploying ML models, running batch processing jobs, scheduling compute-intensive tasks, or serving APIs that require GPU acceleration or dynamic scaling.

17k tokens

How to use it

Copy the folder

Take zhanghandong/makepad-2.0-app-structure from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.