English | 简体中文

mgap (monitor go application profile) is a lightweight desktop pprof monitor built with Wails. It fetches, parses, and visualizes Go runtime profiles in real time, so you don't have to juggle browser tabs and local web servers.

• Live Profile Sampling — continuously sample heap, CPU, allocs, goroutine, block, mutex, and threadcreate profiles.
• Auto Endpoint Detection — just enter a port (e.g. 6060) or a partial URL and the app discovers /debug/pprof endpoints automatically.
• Real-time Charts — top-function timelines powered by ECharts, with configurable time ranges and Top-N filtering.
• Flamegraph View — interactive flamegraph for quick hotspot analysis.
• Import / Export — import existing profile files or export captured snapshots for offline inspection.
• Raw Text View — browser-like plain-text profile output when you need the raw numbers.
• Mock Data Mode — built-in mock data for UI testing without a live Go process.
• Frameless Desktop Shell — custom title bar with drag-to-move, edge-snap maximize / half-screen layout, minimize, and close.
• Cross Platform — Windows, macOS, and Linux support via Wails.

• Go 1.23+
• Wails CLI v2.12+
• Node.js 18+

wails dev

This starts the Go backend and the Vite frontend dev server simultaneously with hot-reload.

wails build

The packaged executable is generated at:

• Windows: build/bin/mgap.exe
• macOS: build/bin/mgap.app
• Linux: build/bin/mgap

package main

import (
	"log"
	"net/http"
	_ "net/http/pprof"
)

func main() {
	log.Println(http.ListenAndServe("localhost:6060", nil))
}

Enter any of the following in the Endpoint input:

6060
localhost:6060
http://localhost:6060/debug/pprof

The app normalizes the input and probes the available profiles.

Click Start to begin periodic sampling. Toggle individual metrics on/off, switch between Flat and Cumulative views, and adjust the Top N filter.

• Drag title bar — move the window.
• Drag to screen top — maximize.
• Drag to left / right screen edge — snap to half-screen.
• Double-click title bar — toggle maximize.

• Profile history is kept in memory and resets when the app restarts.
• Large retained-sample counts can make charts heavier to render.
• This tool is intended for local development and quick inspection, not as a replacement for Prometheus / Grafana.

MIT