惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

月光博客
月光博客
WordPress大学
WordPress大学
博客园 - 三生石上(FineUI控件)
H
Help Net Security
小众软件
小众软件
The Cloudflare Blog
人人都是产品经理
人人都是产品经理
Apple Machine Learning Research
Apple Machine Learning Research
S
SegmentFault 最新的问题
Last Week in AI
Last Week in AI
爱范儿
爱范儿
量子位
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
IT之家
IT之家
博客园 - 【当耐特】
V
Visual Studio Blog
大猫的无限游戏
大猫的无限游戏
博客园_首页
Jina AI
Jina AI
D
Docker
博客园 - 司徒正美
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
Microsoft Security Blog
Microsoft Security Blog
阮一峰的网络日志
阮一峰的网络日志

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
I Made Claude Code Ding When It's Done (And It Changed My...
Anand Rathnas · 2026-06-28 · via DEV Community

Anand Rathnas

This article was originally published on Jo4 Blog.

You know that feeling when you ask Claude Code to refactor a module, switch to Twitter for "just a sec," and come back 12 minutes later to find it's been sitting there waiting for your input for the last 11?

Yeah. That was my Friday evening.

The "Are You Still There?" Problem

I've been using Claude Code as my daily driver while building jo4.io. It's brilliant at crunching through multi-file refactors, running tests, and fixing bugs. But here's the thing - when Claude finishes a task or has a question, it just... sits there. Silently. Like a polite intern who finished their work but doesn't want to interrupt your YouTube rabbit hole.

I needed a way for Claude to tap me on the shoulder. Something that says "Hey, I'm done" or "Hey, I need you" without me obsessively watching the terminal.

The Fix: 5 Lines of JSON

Claude Code has a hook system. I already wrote about PreToolUse hooks for blocking dangerous git commands. Turns out there's a Stop hook that fires every time Claude finishes responding and hands control back to you.

Here's the entire config I added to ~/.claude/settings.json:

{
  "hooks": {
    "Stop": [
      {
        "hooks": [
          {
            "type": "command",
            "command": "afplay /System/Library/Sounds/Funk.aiff",
            "timeout": 5
          }
        ]
      }
    ]
  }
}

That's it. That's the whole thing.

What Happens Now

Every time Claude Code:

  • Finishes a task and waits for my next instruction
  • Runs into something it needs my input on
  • Completes a long test suite run
  • Asks me a clarifying question

...my Mac plays the Funk sound. You know the one - that satisfying little bonk that macOS has shipped since forever.

Why This Is a Game Changer

Before this, my workflow looked like:

1. Give Claude a task
2. Switch to browser
3. Forget about Claude
4. Come back 15 minutes later
5. Realize it's been waiting for 14 of those minutes
6. Feel guilty
7. Repeat

Now it's:

1. Give Claude a task
2. Switch to browser / grab coffee / stretch
3. *bonk*
4. Switch back immediately
5. Continue where we left off

My feedback loop went from "whenever I remember to check" to instant. I'm not exaggerating when I say this cut my average task turnaround in half - not because Claude got faster, but because I stopped being the bottleneck.

Pick Your Sound

macOS ships with a bunch of system sounds. Want something different? Try these:

# List all available system sounds
ls /System/Library/Sounds/

# Some favorites:
afplay /System/Library/Sounds/Glass.aiff    # Subtle, clean
afplay /System/Library/Sounds/Ping.aiff     # Classic notification
afplay /System/Library/Sounds/Hero.aiff     # Triumphant finish
afplay /System/Library/Sounds/Purr.aiff     # Gentle nudge
afplay /System/Library/Sounds/Funk.aiff     # The OG (my pick)

On Linux, you could use paplay, aplay, or even espeak "done" if you want Claude to literally tell you it's finished. On Windows WSL, powershell.exe -c "(New-Object Media.SoundPlayer 'C:\Windows\Media\notify.wav').PlaySync()" works.

Going Further: Conditional Sounds

Want different sounds for different situations? The Stop hook receives JSON on stdin with a last_assistant_message field. You could parse that to play a success sound when tests pass and an error sound when something breaks:

#!/bin/bash
input=$(cat)
message=$(echo "$input" | jq -r '.last_assistant_message // empty')

if echo "$message" | grep -qi "error\|fail\|blocked"; then
  afplay /System/Library/Sounds/Basso.aiff
else
  afplay /System/Library/Sounds/Funk.aiff
fi

Save that as ~/.claude/stop-sound.sh, make it executable, and point your hook at it instead.

The Full Picture

Combined with the PreToolUse hooks I set up earlier, my ~/.claude/settings.json now looks like:

{
  "hooks": {
    "Stop": [
      {
        "hooks": [
          {
            "type": "command",
            "command": "afplay /System/Library/Sounds/Funk.aiff",
            "timeout": 5
          }
        ]
      }
    ],
    "PreToolUse": [
      {
        "matcher": "Bash",
        "hooks": [
          {
            "type": "command",
            "command": "~/.claude/block-git.sh",
            "timeout": 10
          },
          {
            "type": "command",
            "command": "~/.claude/block-cd.sh",
            "timeout": 10
          }
        ]
      }
    ]
  }
}

PreToolUse hooks keep Claude safe. Stop hooks keep me in the loop. Together they make Claude Code feel less like a tool and more like a teammate who knows when to wait and when to nudge.


These are the little things that make your day as a nerdy engineer. A five-line config change, a system sound you've heard a thousand times, and suddenly your entire AI-assisted workflow just clicks. It's not a groundbreaking feature. It's not going to make the front page of Hacker News. But it'll save you dozens of context-switch minutes every single day, and you'll wonder why you didn't set it up sooner.

Have you set up Stop hooks yet? What sound did you pick? Drop a comment - I'm genuinely curious what sounds people gravitate toward. Bonus points if you went the espeak route and have Claude literally talking to you.

Building jo4.io - a modern URL shortener with analytics, bio pages, and team workspaces.