Huashu Design
alchaincyf · Skill
Turns a one-line request in your AI agent into a finished design: a launch animation, a clickable app prototype, an editable slide deck, an infographic or a se…

The Ladybird developers’ web browser, with its own engine built on web standards. It is pre-alpha and for developers, so you build it from source. These notes give the build steps for Linux, macOS, Windows with WSL2, FreeBSD and Android.
Ladybird is an open-source project on GitHub in the Developer category, listed by Mindket. It is free. This is version 2, last updated on . Install it from GitHub as its README says. Get Mindket’s setup notes on your Mindket account, as a zip.
git clone https://github.com/LadybirdBrowser/ladybird.git cd ladybird ./Meta/ladybird.py run
Ladybird is a web browser with its own engine, built on web standards. Its README calls it a truly independent web browser in a pre-alpha state, only suitable for developers. The FAQ says it is not a Chromium shell, not a WebKit port and not a Firefox fork. Many core libraries are inherited from SerenityOS, and the FAQ names Harfbuzz, Skia, simdutf and libcurl among its third-party ones. This entry is about building and running it, since there is no installer to download.
Developers who want to try a new browser engine, read its code or test their sites in it. The README says it is only suitable for developers today. The FAQ roadmap lists an alpha for developers and early adopters on Linux and macOS in 2026, so today you build it rather than install it.
| Industry | Who uses it | How they use it | Example |
|---|---|---|---|
| Browser engineering | C++ developers | Build it, read the engine libraries and work through failing or timing-out WPT tests | A developer opens the timing-out tests on wpt.fyi and isolates one, as the Getting involved page suggests. |
| Web development | Front-end developers | Open their pages in an independent engine and report what renders differently | A front-end developer saves a broken page with SingleFile, cuts it down and posts the reduced case with the issue. |
| Standards work | People who read the specs | Compare a spec algorithm with the code, which the FAQ says follows the spec steps and quotes the spec text in comments | A reader of the HTML spec finds the matching class and checks the quoted steps against the spec. |
| Learning | Students of browser engineering | Read the code beside the book Web Browser Engineering, which the Getting involved page recommends | A student builds the book’s small Python browser first, then opens LibWeb to compare. |
Skip it if you need a daily browser. The README says it is only suitable for developers, and the FAQ says you have to build it from source to try it.
The README says Ladybird runs on Linux, macOS, Windows with WSL2 and many other Unix-like systems. The build page gives steps for Linux, macOS, Windows, FreeBSD and Android.
Ladybird is built with one of two frontends. Android has its own Android UI, and every other platform uses Qt.
Get it saves a zip of Mindket’s setup notes in your downloads folder, with each platform’s steps from the Ladybird build page. The main route is one script, Meta/ladybird.py, once your system’s packages are installed.
| Need | Detail |
|---|---|
| A computer | Linux or macOS, or Windows with WSL2. FreeBSD and Android have their own steps |
| A C++23 compiler | The project’s CI uses gcc-16 and clang-22, and find_compiler.py names the minimum |
| Qt | Version 6.10 or newer. Debian 13 ships Qt 6.8, and configuring against it fails |
| CMake | Version 3.30 or newer in your PATH |
| Rust | A toolchain, which rustup installs |
| Build tools | nasm, ninja and more, listed for each system on the build page |
git clone https://github.com/LadybirdBrowser/ladybird.git cd ladybird ./Meta/ladybird.py run
BUILD_PRESET=Debug ./Meta/ladybird.py run
On macOS with clang from Homebrew, the build page gives this form:
CC=$(brew --prefix llvm)/bin/clang CXX=$(brew --prefix llvm)/bin/clang++ ./Meta/ladybird.py run
For the experimental Clang-CL build on Windows, the build page installs pkg-config for the vcpkg install with Chocolatey, then builds in a VS command prompt:
choco install pkgconfiglite -y py Meta\ladybird.py build
Once the build is done, the run command starts Ladybird. The same script has a few other commands.
| Command | What it does |
|---|---|
| ./Meta/ladybird.py run | Builds a Release version and starts Ladybird |
| BUILD_PRESET=Debug ./Meta/ladybird.py run | Builds a Debug version instead |
| ./Meta/ladybird.py run --no-build | Launches the build you already have, with no configuring or building |
| ./Meta/ladybird.py run <executable_name> | Runs another program, such as the JS REPL or the WebAssembly REPL |
| ./Meta/ladybird.py gdb ladybird | Starts it in gdb |
Put --no-build before the executable name. Anything after the name goes to the executable.
./Meta/ladybird.py run --no-build js --evaluate 'console.log(1 + 1)'
A page looks wrong in Ladybird and fine in another browser. You save it as REDUCTION.html, open that file in Ladybird and check that the same problem happens. Then you cut the file down and reload after each cut. First turn scripting off in Ladybird’s Debug menu to see whether a script causes it. Then remove CSS rules and elements one at a time.
If the page did not come from a SingleFile copy, ISSUES.md says to put a base element into the document, like this, so relative images, style sheets and scripts load:
<base href="https://ladybird.org/">
The smallest file that still shows the problem goes up on a site such as CodePen, JSBin or JSFiddle. Its URL goes into the bug report form.
If you are new. A browser engine is the part that loads a page and draws it. Ladybird writes its own instead of reusing the one in Chromium, WebKit or Firefox. Building it is the first test. Pick your system’s note, install the packages, and run the script. If the build fails, read the troubleshooting note before you ask. The project wants build problems in the build-problems channel on its Discord, not in GitHub issues.
For experts. Each tab runs in its own sandboxed WebContent process. Image decoding and network connections run out of process, in ImageDecoder and RequestServer. The code is split into libraries such as LibWeb, LibJS and LibWasm, and it uses the project’s own AK library instead of the C++ STL. The FAQ says features follow the spec algorithms step by step, with the spec text in comments and an AD-HOC: label on code that maps to no spec step.
For packaging, or for several configurations, a custom build directory avoids the script. The build page gives these lines, and a pool size for machines with little memory. Without the script, you may need to give a suitable C++ compiler with the CMAKE_C_COMPILER and CMAKE_CXX_COMPILER options.
cmake --preset Release -B MyBuildDir cmake --build --preset Release MyBuildDir ninja -C MyBuildDir run-ladybird
cmake --preset Release -B MyBuildDir -DLAGOM_LINK_POOL_SIZE=2
Most of this entry is about a build that runs on your computer. A few things sit outside it.
This kit holds setup notes only. Mindket wrote them. Ladybird is made by the Ladybird developers and is built from source from their repository: https://github.com/LadybirdBrowser/ladybird
These notes are not made or endorsed by the Ladybird project or the Ladybird Browser Initiative.
License: BSD 2-Clause. Source: https://github.com/LadybirdBrowser/ladybird/blob/master/LICENSE (its copyright line reads "Copyright (c) 2018-present, the Ladybird developers."). The README says "Ladybird is licensed under a 2-clause BSD license."
Ladybird is a web browser with its own engine, built on web standards. Its README calls it "a truly independent web browser" and says it is in a pre-alpha state, "only suitable for use by developers". There is no installer and no download. Its FAQ says you have to build it from the sources yourself to try it. This kit gathers the build and run steps its docs give for each platform.
A multi-process design: a main UI process, several WebContent renderer processes, an ImageDecoder process and a RequestServer process. Each tab has its own renderer process, sandboxed from the rest of the system. Image decoding and network connections happen out of process. At the moment many core library components are inherited from SerenityOS: LibWeb (the web rendering engine), LibJS (the JavaScript engine), LibWasm, LibCrypto and LibTLS, LibHTTP, LibGfx, LibUnicode, LibMedia, LibCore and LibIPC.
The route the project's home page gives, "on many development machines", once the prerequisites for your system are installed:
git clone https://github.com/LadybirdBrowser/ladybird.git
cd ladybird
./Meta/ladybird.py run
All three lines are on the project's home page (https://ladybird.org), and the build instructions call the last one the simplest way to build and run. A release build is the default.
The project's CONTRIBUTING.md says Ladybird is developed by a group of project maintainers who introduce all changes to the codebase, and that it does not accept pull requests from non-maintainers. Bug reports, reductions, website testing, standards discussion, security reports and technical feedback all help. See setup/filing-issues.md.
README.md 4 KB setup/android.md 2 KB setup/building-and-running.md 4 KB setup/filing-issues.md 5 KB setup/freebsd.md 1 KB setup/linux.md 6 KB setup/macos.md 3 KB setup/troubleshooting.md 4 KB setup/windows.md 1 KB
Updated by Mindket to the current version of this kit.
Posts and comments that name this project on Hacker News, Bluesky, X, Reddit, Facebook and YouTube, found by a weekly search and on Trendshift. Mindket doesn’t write or check them. Open one to read it where it was posted.
alchaincyf · Skill
Turns a one-line request in your AI agent into a finished design: a launch animation, a clickable app prototype, an editable slide deck, an infographic or a se…
op7418 · Skill
Skill that makes single-file HTML slide decks, slide illustrations and platform covers in two visual styles, with a presenter mode for rehearsal. For speakers…
alirezarezvani · Skill
Adds 411 ready-made skills for Claude Code and other coding agents, grouped by role from engineering and product to marketing, finance and compliance. Installs…
block · Kit
Self-hostable workspace where people and AI agents share the same rooms, built on a Nostr relay that logs every action as a signed event. For teams that want t…
backnotprop · Kit
Opens your AI coding agent's plans, specs and code changes in a browser so you can annotate or comment on them and send the feedback straight back to the agent.
alibaba · Kit
Reviews Git changes with an AI agent that reads files and searches the codebase, then writes comments tied to exact lines. It suits developers who want automat…