Welcome to The Weekly Five - your curated list of 5
exceptional open source projects I discovered this week.

The Weekly Five

This week's lineup is all about reclaiming control over the tools you use every day. Whether you are wrangling AI agents in your terminal, taming Windows bloat, building cross-platform apps from a single codebase, crafting buttery animations, or exploring a browser built from scratch: these five projects put you back in the driver's seat.

Top takeaways

  • Terminal-native AI orchestration is maturing fast, and herdr shows what "tmux for agents" can look like in practice

  • Cross-platform development keeps getting more accessible, with Quasar letting you ship to web, mobile, and desktop from one Vue.js codebase

  • The browser wars have a new independent contender: Ladybird is building a novel engine from the ground up, no Chromium or WebKit in sight

Who this issue is for

Developers and power users who want practical tools that simplify system management, streamline builds, and unlock smoother workflows across platforms.

1. Ladybird

Why this made the cut: A truly independent browser engine is rare, and Ladybird represents a serious attempt to break the Chromium/WebKit duopoly.

Why it matters

Modern browsers share a handful of underlying engines, which limits diversity and innovation on the web. Ladybird is building a novel engine from scratch, aiming to give developers (and eventually everyone) a genuinely independent alternative. Even in pre-alpha, it is a signal that browser competition is not dead.

Key features

  • Completely independent rendering engine (LibWeb), not a fork of Blink, WebKit, or Gecko

  • Standards-focused architecture designed for long-term maintainability

  • Open development with over 1,500 contributors already involved

How to use

  1. On Debian/Ubuntu, install the build dependencies: sudo apt install build-essential cmake ninja-build nasm autoconf automake libtool pkg-config qt6-base-dev qt6-tools-dev-tools libgl1-mesa-dev libdrm-dev python3-venv

  2. Make sure you have a C++23-capable compiler (GCC 14+ or Clang 20+); on older Ubuntu, add the toolchain PPA and install g++-14

  3. Clone the repo and build plus launch in one step: git clone https://github.com/LadybirdBrowser/ladybird && cd ladybird && ./Meta/ladybird.py run

Expect rough edges: this is pre-alpha software meant for experimentation and contribution, not daily browsing. 🔗 View on GitHub | GitHub stars: 64,678

Learning resources

  1. Ladybird Browser: First Impressions & Easy Installation Guide (2024) - Habr - Walks through installing dependencies and building Ladybird from source on Ubuntu.

2. WinUtil

Why this made the cut: A single PowerShell command that handles installs, debloating, and system tweaks saves hours on every fresh Windows setup.

Why it matters

Setting up a new Windows machine typically means clicking through dozens of menus, disabling telemetry, removing preinstalled apps, and hunting for drivers. WinUtil consolidates all of this into one script you can run immediately after install, giving you a cleaner, faster system without the manual busywork.

Key features

  • Batch install popular programs via a curated list

  • One-click debloat tweaks to remove unwanted Windows components

  • Config options for troubleshooting and Windows Update management

How to use

  1. Right-click the Start menu and open Windows PowerShell (Admin) or Terminal (Admin)

  2. Run the install command from the repo README to launch the utility

  3. Select the tweaks, programs, or fixes you want, then apply Run it fresh on every new Windows install for a consistent, bloat-free baseline. 🔗 View on GitHub | GitHub stars: 59,099

3. herdr

Why this made the cut: Managing multiple AI coding agents in one terminal session finally has a purpose-built solution.

Why it matters

If you run Claude Code, Codex, or other AI agents, you know the pain of juggling multiple terminal sessions. herdr is a Rust-based multiplexer designed specifically for agent orchestration, with built-in awareness of agent state. Think of it as tmux, but tailored for the AI-assisted coding workflow.

Key features

  • Agent-aware pane management that tracks the state of each running agent

  • Keybindings and UI designed for fast context switching

  • SSH remote usage support for distributed setups

How to use

  1. Install herdr from the latest release (prebuilt binaries available)

  2. Launch herdr and open panes for each AI agent (Claude Code, Codex, etc.)

  3. Use built-in keybindings to switch between agents, monitor their state, and manage your workspace Community tip: Pair herdr with your existing dotfiles for a seamless terminal experience. 🔗 View on GitHub | GitHub stars: 21,374

Learning resources

  1. Herdr: the Tmux for AI Agents - YouTube - Hands-on tour of running several coding agents side by side in herdr panes.

  2. Herdr: Terminal Multiplexer with Built-in AI Agent State Awareness - Better Stack - Explains herdr's agent state tracking and how it differs from plain tmux.

4. React Native Reanimated

Why this made the cut: Smooth, performant animations are table stakes for mobile UX, and Reanimated makes them accessible without dropping into native code.

Why it matters

React Native's built-in Animated library can feel limiting when you need complex, interruptible, or gesture-driven animations. Reanimated reimplements the animation primitives to run on the UI thread, unlocking 60fps performance and more expressive motion design.

Key features

  • Shared values and animated styles for declarative, reactive animations

  • Worklets that run directly on the UI thread (no JS bridge bottleneck)

  • Composable timing functions: withSequence, withDelay, withRepeat, and more

How to use

  1. Install via npm or yarn and follow the setup steps for your platform (iOS/Android)

  2. Define shared values and animated styles in your components

  3. Use worklets and timing functions to orchestrate complex animations Usage pattern: Combine Reanimated with gesture handlers for swipe-to-dismiss, drag-and-drop, and other interactive effects. 🔗 View on GitHub | GitHub stars: 10,935

Learning resources

  1. Introduction to React Native Animations with Reanimated - YouTube - Beginner walkthrough of shared values and animated styles in a real component.

  2. Introduction to Reanimated - Reactiive - Written primer on worklets and the UI-thread animation model.

5. Quasar

Why this made the cut: One codebase for SPAs, PWAs, mobile apps, Electron apps, and browser extensions is a rare promise that Quasar actually delivers on.

Why it matters

Cross-platform development often means juggling multiple frameworks or accepting lowest-common-denominator UX. Quasar leverages Vue.js to let you build for web, iOS, Android, desktop (Electron), and even browser extensions from a single codebase, with a rich component library and sensible defaults.

Key features

  • Material Design components out of the box

  • First-class support for SSR, SSG, PWA, and hybrid mobile builds

  • CLI tooling that scaffolds and configures projects for any target platform

How to use

  1. Install the Quasar CLI globally via npm

  2. Create a new project and select your target platforms (SPA, PWA, Electron, etc.)

  3. Build and deploy to each platform using the same source code Community tip: Start with a PWA target to validate your UI, then expand to native mobile or Electron as needed. 🔗 View on GitHub | GitHub stars: 27,206

Learning resources

  1. Master Quasar: Build Powerful Cross-Platform Apps - Vue School - Course intro covering CLI setup and building for multiple target platforms.

  2. Uncover Quasar Framework (Simplified Beginner's Guide) - BytesTechnoLab - Beginner overview of Quasar's component library and project structure.

If you only try one

herdr is the single best starting point this week. If you are already using AI coding agents (or curious about them), herdr solves a real, immediate pain point: managing multiple agents in one terminal without losing track of what each is doing. It is lightweight, installs in seconds, and the workflow benefits are obvious from the first session.

If you are doing Open Source I have a good news for you, I work at CodeRabbit which is an AI review tool and its free for Open Source, please reach out to me on X or LinkedIn or just send an email on [email protected] if you need help on adopting CodeRabbit.

You can visit our portal below to create a new account and connect your repository and start reviewing your code.

Keep Reading