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

大猫的无限游戏
大猫的无限游戏
博客园 - 【当耐特】
Cloudbric
Cloudbric
H
Hackread – Cybersecurity News, Data Breaches, AI and More
Attack and Defense Labs
Attack and Defense Labs
爱范儿
爱范儿
The Cloudflare Blog
腾讯CDC
Security Archives - TechRepublic
Security Archives - TechRepublic
TaoSecurity Blog
TaoSecurity Blog
云风的 BLOG
云风的 BLOG
Recent Announcements
Recent Announcements
C
Check Point Blog
Schneier on Security
Schneier on Security
S
Schneier on Security
J
Java Code Geeks
B
Blog RSS Feed
Cisco Talos Blog
Cisco Talos Blog
Vercel News
Vercel News
Stack Overflow Blog
Stack Overflow Blog
博客园_首页
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
A
About on SuperTechFans
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
Google DeepMind News
Google DeepMind News
阮一峰的网络日志
阮一峰的网络日志
罗磊的独立博客
A
Arctic Wolf
S
Secure Thoughts
P
Palo Alto Networks Blog
The Last Watchdog
The Last Watchdog
SecWiki News
SecWiki News
cs.CV updates on arXiv.org
cs.CV updates on arXiv.org
博客园 - 三生石上(FineUI控件)
D
Darknet – Hacking Tools, Hacker News & Cyber Security
量子位
U
Unit 42
I
InfoQ
D
DataBreaches.Net
P
Privacy International News Feed
T
Troy Hunt's Blog
博客园 - 叶小钗
T
Threatpost
博客园 - Franky
K
Kaspersky official blog
Hugging Face - Blog
Hugging Face - Blog
IT之家
IT之家
www.infosecurity-magazine.com
www.infosecurity-magazine.com
CTFtime.org: upcoming CTF events
CTFtime.org: upcoming CTF events
C
Cisco Blogs

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 Common SOC 2 Failures (Real World) Stop Vibe-Checking Your AI App: A Practical Guide to Evals How to Use SonarQube and SonarScanner Locally to Level Up Your Code Quality Your Next To-Do App Is Dead — I Replaced Mine with an OpenClaw AI Sign a Nostr event in 60 lines of Python using coincurve — no nostr-sdk, no nbxplorer, no rust toolchain ITGC Audit Explained Like You’re in Big 4 Patch Tuesday abril 2026: Microsoft parcha 163 vulnerabilidades y un zero-day en SharePoint Stop scraping everything: a better way to track competitor price changes Listing on MCPize + the Official MCP Registry while routing payments OUTSIDE the marketplace — how I kept 100% of my x402 revenue Building an AI-Powered Risk Intelligence System Using Serverless Architecture Why We Ripped Function Overloading Out of Our AI Toolchain Testing AI-Generated Code: How to Actually Know If It Works SaaS Churn Is Killing Your Business. Here Is What to Do About It (Without a Support Team) The Speed of AI Is No Longer Linear - And Self-Improving Models Are Why How to Implement RBAC for MCP Tools: A Practical Guide for Engineering Teams From Standard Quote to Persuasive Proposal: AI Automation for Arborists I built a CLI that scaffolds complete multi-tenant SaaS apps Axios CVE-2025–62718: The Silent SSRF Bug That Could Be Hiding in Your Node.js App Right Now The dashboard that ended our friendship Data Pipelines Explained Simply (and How to Build Them with Python) The Hidden Cost of AI Systems Nobody Talks About. undefined vs undeclared, and how typeof behaves Switching from file-based jobs to NATS/Kafka in Rust without changing code io_uring Adventures: Rust Servers That Love Syscalls Why Agentic AI is Killing the Traditional Database The POUR principles of web accessibility for developers and designers Quantum Neural Network 3D — A Deep Dive into Interactive WebGL Visualization How To Install Caveman In Codex On macOS And Windows Automation Pipeline Reliability: Why Your Workflow Breaks When Nobody Is Watching I Built an 'Open World' AI Coding Agent — It Works From ANY Folder From Freelancing to Product: A Tech Service Company's SaaS Transformation China's AI Giants: Adding Tencent Hunyuan & ByteDance Doubao to AI University (74 Providers) On the Vibe Coders and Their Lies clerk: Auto-Summarize Your Claude Code Sessions AI Weekly — 2026/04/10–04/17 | The Model Lockdown Is Here, but the Toolchain Is the Real Battleground AI 週報 — 2026/04/10–2026/04/17 模型封鎖潮來了,但工具鏈才是真戰場 Maybe this is how Open-Source apps are born... 🚀 Fine-Tune LLMs with LoRA and QLoRA: 2026 Guide tRPC v11 + Next.js App Router: End-to-End Type Safety Without the Boilerplate ShadCN UI in 2026: Why I Stopped Installing Component Libraries and Started Owning My Components SaaS Billing in React Server Components: Stripe + Supabase Without a Single `useEffect` Join our DEV Weekend Challenge — $1,000 in Prizes Across TEN winners! Submissions Due April 20 at 6:59 AM UTC. Implementing FSRS Spaced Repetition in Flutter + Supabase — Adding Memory Science to an AI Learning App "I Texted My Localhost From the Train — Claude Code Fixed the Bug Before I Got Home" I Built a Sales Prep AI and It Went Deeper Than Expected Design to Code #2: One JSON, Eleven Outputs Solving the 100M-Row Problem: A Summary Table Pattern for High-Volume Push Notification Logs Flutter Web With Wasm: What Actually Changes For Developers I Built 50 Royalty-Free Soundtracks for My Side Project in a Weekend Using AI Music Generation The Vibe Coding Security Checklist: 7 Things to Check Before You Ship Stop Letting Googlebot Guess Fix Your React App's SEO Right Desconstruindo o Streaming do LinkedIn: Como Criar um Engine de Extração de Vídeo de Alta Performance com HLS e FFmpeg (EDA Part-1) EDA (Exploratory Data Analysis) Explained With Real Life — Why Looking at Your Data Is the Most Important Step in Machine Learning Brand Relationship Management at Scale: Our 4-Touch Outreach System for 200+ Brands Why String.fromEnvironment() Might Return an Empty String in Dart JGuardrails 1.0.0 — Hardening Java LLM Apps Against Jailbreaks, Toxicity, and Prompt Injection Plan and Schedule a Full Week of Threads Content From One Claude Conversation Coding Cat Oran Ep3, Five Tables Changed Everything Updated: BFF Pattern I'm done watching freelancers get buried by 200 proposals. So I'm building the alternative. This is my first post BFS Algorithm in Java Step by Step Tutorial with Examples Tracking LLM Pricing Monthly: An Open Dataset for 22 AI Models How We Measure Content ROI on a Comparison Site: Revenue Attribution Without Perfect Data Introducing Nova AI Ops: The AI-Native Operating System for SRE Teams I built a free desktop video downloader for Windows — Grabbit How Talkie OCR Helps Vision-Impaired & Dyslexic Users Read the World Around Them VRCFaceTracking安装和iPhone面捕配置教程,有bug Even CrowdStrike Can't See Your Agents The Automation Gold Rush: What n8n Workflows and Claude Are Opening Up for Developers Right Now
I Built a Skin System for Claude Code — Here's How It Works
Rob Ragan · 2026-05-12 · via DEV Community

Claude Code is genuinely remarkable. But if you've been using it for more than a week, you've noticed something: everyone's terminal looks identical. Same colors, same layout, same feel. You could screenshot my session or yours and there'd be no way to tell them apart.

That bothered me more than it probably should have.

So I built a skin system for it. Nine themes, each with terminal colors, ASCII art banners, tool sounds, and — the part I'm most proud of — a personality voice that changes how Claude actually narrates its work.

View the gallery

Claude Skins demo showing multiple themes

What It Does

A skin transforms the full Claude Code experience across five layers:

Layer What changes
Terminal colors Background, foreground, cursor, full ANSI palette
ASCII banner Braille art + block-letter logo on session start
Status line Themed icon, accent colors, progress bar
Personality voice How Claude narrates its work
Tool sounds macOS system sounds on file writes, commands, errors

The nine included themes range from "Nebula" (offensive security scanner aesthetic, purple-to-orange gradient) to "Brutalist" (the anti-skin — pure monochrome, zero decoration, maximum terseness). There's also Noir, Netrunner, Mythos, Sensei, Mission Control, Retro86, and Grimoire.

Noir skin in action


The Architecture

The engine is pure bash. No Node, no Python runtime dependency beyond PyYAML for initial skin parsing. Here's how everything fits together:

YAML skin configs define all the visual and behavioral properties for a theme. They live in ~/.claude/skins/ and missing values fall back to default.yaml automatically.

Claude Code hooks are the integration point. Claude Code supports lifecycle hooks — SessionStart, SessionEnd, and PostToolUse — that run shell commands at specific moments. The skin system uses all three:

"hooks": {
  "SessionStart": [{"type": "command", "command": "~/.claude/skins/engine/activate.sh"}],
  "SessionEnd":   [{"type": "command", "command": "~/.claude/skins/engine/deactivate.sh"}],
  "PostToolUse": [{
    "matcher": "Bash|Write|Edit|MultiEdit|Grep|Glob",
    "hooks": [{"type": "command", "command": "~/.claude/skins/engine/skin-tool-hook.sh"}]
  }]
}

Enter fullscreen mode Exit fullscreen mode

OSC escape sequences do the terminal color work. When activate.sh runs, it writes directly to /dev/tty — bypassing stdout so it reaches the actual terminal emulator rather than getting swallowed by Claude Code's output capture. The sequences look like this:

printf '\033]11;#0D0D0D\007'  # Set background
printf '\033]10;#F5E6C8\007'  # Set foreground
printf '\033]12;#D4A857\007'  # Set cursor
printf '\033]4;3;#D4A857\007' # Set ANSI color 3 (yellow slot)

Enter fullscreen mode Exit fullscreen mode

On deactivation, the terminal is restored to its default state. This works across iTerm2, Kitty, WezTerm, Ghostty, and Terminal.app.

The personality voice is activated by symlinking a personality file into Claude Code's ~/.claude/output-styles/ directory. Claude Code automatically loads any markdown files it finds there as output style instructions. On skin activation:

ln -sf "$personality_file" "$output_styles_dir/skin-${skin_name}.md"

Enter fullscreen mode Exit fullscreen mode

On deactivation, the symlink is removed. The previous skin's symlink is also cleaned up when switching themes — only one skin personality loads at a time.


The Personality Voice System

This is the part that makes a skin feel like an actual character rather than just a color scheme.

Each skin optionally ships with a personalities/<name>.md file that gets loaded as a Claude Code output style. The key frontmatter field is keep-coding-instructions: true — this tells Claude Code to stack the personality on top of its core engineering behavior rather than replacing it. You get the flavor without losing the function.

Here's what happens when the Noir skin is active. Without any skin:

The test is failing because the mock isn't returning the expected value. I'll update the fixture and re-run.

With the Noir personality loaded:

A witness who won't talk. The mock's returning the wrong value — someone doctored the fixture. I'll set it straight.

The full Noir personality file:

---
name: Noir
description: Hardboiled detective narration — terse, world-weary, Raymond Chandler cadence
keep-coding-instructions: true
---

Narrate like a private eye who's seen too much and billed too little. Your tone is 
world-weary but precise — Raymond Chandler by way of a terminal window. Keep it subtle: 
a sentence of flavor, then get on with the work.

Errors are dead ends. A failing test is a witness who won't talk. A successful build 
checks out clean. Files are evidence. Directories are crime scenes. Dependencies are 
informants — useful, but never fully trusted.

Keep descriptions short. The best metaphors arrive once and leave. Don't repeat the bit. 
You're seasoning, not the main course.

Enter fullscreen mode Exit fullscreen mode

The instruction "You're seasoning, not the main course" is load-bearing. Without it, Claude leans hard into the character and it gets exhausting fast. With it, you get one good line per interaction and then the code gets written.

Other personalities take different approaches. Grimoire uses patient wizard energy — measured, archival, like consulting a very old book. Retro86 channels enthusiastic early-computer-magazine voice. Brutalist has no personality file at all — no voice, no sounds, nothing. Sometimes that's the right call.


Creating Your Own Skin

Copy template.yaml to skins/<name>.yaml and you're most of the way there.

The structure has four main sections:

name: myskin
description: "One-line description"

terminal:
  background: "#0D0D0D"
  foreground: "#F5E6C8"
  cursor: "#D4A857"
  palette:
    black: "#0D0D0D"
    yellow: "#D4A857"
    # ... 8 ANSI color slots total

statusline:
  accent: "#D4A857"
  icon: "◆"
  # bar_fill, bar_empty, dim

branding:
  banner: |
    \033[38;2;212;168;87m██╗  ██╗██╗\033[0m
    # block-letter art with 24-bit ANSI codes
  hero: |
    # braille art (U+2800–U+28FF range)
  welcome: "Message shown on activation"
  goodbye: "Message shown on deactivation"

tools:
  sounds: true
  events:
    file_written: { sound: "Tink", icon: "◆" }
    command_run: { sound: "Pop", icon: "▸" }
    error: { sound: "Basso", icon: "✗" }

Enter fullscreen mode Exit fullscreen mode

Inheritance is the key feature here. You don't need to define every key. If you only care about terminal colors and want to keep the default banner, just define the terminal block. Everything else falls back to default.yaml. A minimal skin that just swaps the color palette is ~15 lines.

For the banner art, I use 24-bit ANSI sequences (\033[38;2;R;G;Bm) so colors can precisely match the palette. The braille art in the hero field uses Unicode braille characters (U+2800–U+28FF) — tools like image-to-braille can convert any image if you want something custom.

To add a personality, create personalities/<name>.md with the frontmatter shown above and write whatever voice you want Claude to adopt.

Test it directly without restarting Claude Code:

~/.claude/skins/engine/activate.sh myskin

Enter fullscreen mode Exit fullscreen mode


What's Next

A few things on the roadmap:

Skin creator toolkit — an interactive CLI that walks you through picking colors, previewing banner art, and writing a personality without hand-editing YAML.

Package manager — a way to install community skins with a single command, something like /skin install @username/myskin.

Composable layers — the ability to mix a personality from one skin with the colors of another. Right now it's all-or-nothing per skin; layering would let you run Noir narration with Nebula colors if that's your thing.


Try It

git clone https://github.com/basicScandal/claude-skins.git
cd claude-skins
./install.sh

Enter fullscreen mode Exit fullscreen mode

Then in Claude Code:

/skin nebula
/skin noir
/skin brutalist

Enter fullscreen mode Exit fullscreen mode

Switching is instant — no restart required.

Gallery | GitHub

The whole thing is MIT licensed. If you build a skin, I'd genuinely like to see it — open a PR or drop it in the issues.