Skip to content
Get packs

Guide · Agent skill

How to use the Desktop Brand Builder agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

Desktop Brand Builder is an agent skill from the QwenLM/qwen-code repository that turns the Qwen Code Tauri desktop app into a client with your own name, icons and startup screen. It is aimed at teams and developers who want a white-label desktop installer, not at people who just want to use Qwen Code. After this guide you will know what the skill needs, how to install it, how to ask the agent for a build, and what to check before you hand an installer to anyone.

Desktop Brand BuilderQwenLM/qwen-code · Apache-2.0

Quick start

  1. 1Install the skill. Unzip the downloaded ZIP into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills.
  2. 2Confirm prerequisites. Make sure Git, Node.js 18 or newer, npm, and the Rust and Tauri build prerequisites are installed on your machine.
  3. 3Prepare your brand inputs. Choose a brandId using lowercase letters, digits and hyphens, and a square PNG logo of at least 1024 pixels. Add a website if you have one.
  4. 4Ask the agent for a build. Describe the brand and target platform in plain language, for example macOS DMG or Windows EXE, and provide the brandId and logo path.
  5. 5Let the agent clone and patch. The agent clones the repository into a separate build folder and runs the bundled brand-create.mjs script rather than editing files by hand.
  6. 6Package and verify. The agent runs the packaging step, then checks that the artifact exists, computes its SHA-256 and reports the results.

What an agent skill is and what this one does

An agent skill is a folder containing a SKILL.md instruction file and optional scripts. When installed, the agent reads the instructions when your request matches the skill's description, then follows the steps it contains.

This skill takes a short brand spec: an ID such as acme-ai, a local logo file, and optionally a website. The agent clones the repository into a separate build folder, runs a bundled Node script that patches the app's config, icons and startup screen, and then packages an installer for the platform you choose. The skill tells the agent not to hand-edit those files and to report honestly which installers were actually produced.

Install the skill

Download the skill ZIP from this page and unzip it. For a personal install, place the folder in ~/.claude/skills/. For a single project, place it in that project's .claude/skills/ folder.

If you use the Claude apps instead, upload the ZIP under Settings → Capabilities → Skills.

  • Personal: ~/.claude/skills/
  • Project: .claude/skills/ inside your project
  • Claude apps: Settings → Capabilities → Skills, then upload the ZIP

Check what you need before building

The build runs on your own machine and needs several tools. The skill runs scripts and uses the network, so read the requirements first.

  • Git
  • Node.js 18 or newer, and npm
  • Rust toolchain and the Tauri build prerequisites for packaging
  • Network access to clone github.com/QwenLM/qwen-code and fetch npm packages
  • Optional: your own updater signing key and update feed if in-app updates are wanted

Trigger the skill with a natural request

You do not need a special command. Ask in plain language for a custom, white-label or rebranded desktop client, installer, DMG, EXE, AppImage or deb built on top of the desktop package. The agent should then ask once for anything missing.

The required inputs are a brandId and a logo. The brandId may use only lowercase letters, digits and hyphens, and the logo should be a square PNG of at least 1024 pixels. The website is optional. The skill says not to ask for app name, app ID or updater endpoints unless you want to override them.

  • Example request: Make a branded Qwen Code desktop installer for acme-ai using /absolute/path/to/logo.png, website https://acme.ai, for macOS.
  • Example request: Build a white-label DMG for brand acme-ai with my logo file.

What the build produces and how to verify it

The agent clones the repository into a brand-builds folder, writes a temporary brand.json, installs dependencies, runs the bundled brand-create.mjs script, then runs the packaging commands. Artifacts appear under packages/desktop/src-tauri/target/release/bundle/ for the host platform.

The skill asks the agent to confirm that the artifact exists, compute a SHA-256 checksum, run hdiutil verify on DMG files on macOS, and report the artifact path, checksum, app name, app ID and build directory. Expect the agent to name only the installers that actually exist.

  • Derived defaults: App name Acme AI, artifact prefix Acme-AI, and app ID ai.acme.desktop when the website is https://acme.ai
  • Builds are unsigned by default
  • Updater endpoints are empty by default, so branded builds do not poll the official Qwen Code feed

Signing and in-app updates

Branded builds are unsigned by default. The official Apple, Windows and updater signing secrets belong to the upstream Qwen Code releases and must not be reused. If you need signed releases or in-app updates, you must set up your own credentials and your own update feed.

The skill documents a Tauri signer command for creating an updater key pair. The private key stays in your build environment, and the public key goes into brand.json as updaterPubkey.

npx @tauri-apps/cli signer generate -w ~/.tauri/my-brand.key

What people use it for

  • Give a partner company a Qwen Code based desktop client carrying their logo and name
  • Maintain an internal white-label desktop app for a team with a repeatable build
  • Produce a DMG on macOS or an AppImage or deb on Linux for a branded pilot group
  • Test how a rebranded shell looks by generating an unsigned build for review

Frequently asked questions

How do I install the desktop-brand-builder skill?

Download the skill ZIP from this page and unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps, upload the ZIP under Settings → Capabilities → Skills instead.

What do I need to run desktop-brand-builder?

You need Git, Node.js 18 or newer, npm, and the Rust toolchain with Tauri build prerequisites for packaging. The build also needs network access to clone the Qwen Code repository and fetch npm packages.

Is desktop-brand-builder free and what is its licence?

The item is listed under the Apache-2.0 licence. The source does not describe pricing beyond that, so check the repository licence file for the terms that apply to your use.

Why are the branded builds unsigned?

The skill says branded builds are unsigned by default because the official Apple, Windows and updater signing secrets belong to the upstream Qwen Code releases. To get signed or auto-updating builds you need your own signing keys and your own update feed.

Can I use this instead of building Qwen Code myself?

This skill automates the rebranding and packaging steps on top of packages/desktop. If you only want to use Qwen Code day to day, you do not need it; it is a build tool for teams shipping their own desktop client.

Why does my build fail or the app won't launch?

The skill warns that building for a different platform requires re-running build:runtime with QWEN_DESKTOP_TARGET set before each tauri build --target, otherwise the Node binary has the wrong architecture and fails with an exec format error. It also refuses to run on a shell whose productName is already branded, so start from a fresh clone.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Desktop Brand Builder suits teams and developers who maintain a white-label Qwen Code desktop client and want a repeatable build. It is a poor fit for ordinary users, and for anyone who needs signed or auto-updating releases without setting up their own keys and update server.

Download the skill