mrkeyoor.com_
Wed 16 Sept 18:57 UTC

Best Open-Source Dev Tools

Editors, terminals, and utilities that earn a place in a daily workflow. Ranked by our usability scores across setup ease, docs, community health, and maturity — 532 tools reviewed in full.

01

fzf

junegunn/fzf · ★ 83k113/7d · Go

fzf is an interactive fuzzy finder for the terminal. It reads lines, lets you narrow them with a forgiving query, and returns the selected entry, which turns shell output into a keyboard-driven picker.

SetupDocsCommunityMaturity

Verdict Our fzf run installed 19 packages, built in 15 seconds, and passed all 4 tested Go packages, making it an unusually low-risk addition to a terminal workflow. Use it when selection is the missing step between Unix commands, especially if previews or live reloads can replace a custom script.

Full review →
02

lazygit

jesseduffield/lazygit · ★ 82k191/7d · Go

Lazygit is a keyboard-driven terminal interface for Git. It puts files, branches, commits, stashes, diffs, rebases, worktrees, and common remote actions into panels, so developers can inspect and change a repository without recalling every command flag.

SetupDocsCommunityMaturity

Verdict Our lazygit build took 38 seconds and all 33 detected tests passed, making it the cleanest source run in this group. Use it if you already understand Git and want staging, rebase, worktree, and history tasks kept in one terminal view.

Full review →
03

ripgrep

BurntSushi/ripgrep · ★ 68k167/7d · Rust

ripgrep is a command-line search tool for finding text and regular-expression matches across directories. It searches recursively, respects ignore files by default, skips hidden and binary files, and works on Windows, macOS, and Linux.

SetupDocsCommunityMaturity

Verdict Our ripgrep run installed 42 packages in 9 seconds, built in 19 seconds, and passed all 450 tests in 11 seconds. It is the sensible default for interactive source-tree search on a developer machine, provided you understand that ignored, hidden, and binary files are excluded unless requested.

Full review →
04

bat

sharkdp/bat · ★ 60k69/7d · Rust

bat is a command-line file viewer that adds syntax coloring, line numbers, Git change markers, and automatic paging to the job normally handled by `cat`. It makes source files easier to scan in a terminal while falling back to plain output when another program consumes the result.

SetupDocsCommunityMaturity

Verdict Our bat run passed all 443 tests after a 22-second install and 76-second build, making it an easy recommendation for developers who read files in a terminal. Install a packaged binary and keep `cat` available for bare, universal behavior.

Full review →
05

requests

psf/requests · ★ 54k14/7d · Python

Requests is a synchronous Python library for making HTTP/1.1 calls without handling connection details by hand. It covers URL parameters, forms, JSON, authentication, sessions, cookies, proxies, TLS verification, uploads, and streamed downloads through a small API.

SetupDocsCommunityMaturity

Verdict Our Requests run installed 54 packages in 37 seconds, built in 6 seconds, and reported 619 passed, 0 failed, and 15 skipped of 619. Use it as the default choice for straightforward synchronous HTTP/1.1 in Python, especially when compatibility and a stable interface matter.

Full review →
06

prettier

prettier/prettier · ★ 52k48/7d · JavaScript

Prettier parses source code and prints it again using a fixed set of formatting rules, replacing team debates about spaces, wrapping, quotes, and line breaks with one repeatable output. It supports JavaScript, TypeScript, CSS, HTML, JSON, Markdown, YAML, GraphQL, and several related formats in core, with plugins covering more languages.

SetupDocsCommunityMaturity

Verdict Our Prettier checkout passed 35,265 of 35,270 tests with 0 failures after a 38-second install and 23-second build, so it remains the formatter I would choose first for a web team that accepts its house style. Pin the exact version, introduce the initial rewrite as its own change, and enforce `--check` in CI.

Full review →
07

ruff

astral-sh/ruff · ★ 50k89/7d · Rust

Ruff is a Python linter and formatter written in Rust. It combines style checks, import sorting, code upgrades, and formatting in one executable, reducing the number of tools a Python project must configure.

SetupDocsCommunityMaturity

Verdict Most modern Python projects should try Ruff, starting with linting and adding its formatter if the documented differences are acceptable. It can replace a surprising amount of routine tooling, but it should sit beside a type checker and any unsupported specialist checks.

Full review →
08

cobra

spf13/cobra · ★ 45k12/7d · Go

Cobra is a Go library for building command-line programs with nested commands, flags, help text, suggestions, documentation, and shell completion. It handles the interface around a CLI so application code can focus on what each command does.

SetupDocsCommunityMaturity

Verdict Our Cobra checkout built in 15 seconds and all 2 measured test packages passed in 9 seconds after installing only 7 packages. It remains the safe default for a Go CLI with nested commands, generated help, and several completion targets.

Full review →
09

fd

sharkdp/fd · ★ 44k · Rust

fd is a Rust command-line tool for finding files and directories with shorter syntax and project-friendly defaults. It searches names with regular expressions, respects ignore files, skips hidden entries by default, and can run commands over its results.

SetupDocsCommunityMaturity

Verdict Our fd checkout installed in 8 seconds and passed all 538 tests in 12 seconds, making it the easiest recommendation in this batch for interactive project search. Use it as your everyday finder, especially in Git repositories.

Full review →
10

tokio

tokio-rs/tokio · ★ 33k62/7d · Rust

Tokio is a Rust runtime for programs that wait on networks, timers, processes, and other input or output without blocking a thread for each job. It supplies the scheduler, operating-system event loop, sockets, channels, and timing tools that async Rust code needs to run.

SetupDocsCommunityMaturity

Verdict Our Tokio run built in 35 seconds and passed all 4,148 tests in 364 seconds, the cleanest result in this review set. Use it by default for a Rust network application unless a measured workload calls for a smaller or thread-per-core runtime.

Full review →
11

eslint

eslint/eslint · ★ 28k12/7d · JavaScript

ESLint reads JavaScript, applies a configurable set of rules to its syntax tree, and reports code patterns your team wants to catch. It can also fix some findings, while plugins and alternate parsers extend it to frameworks and languages such as React and TypeScript.

SetupDocsCommunityMaturity

Verdict Our ESLint run installed 1,067 packages in 133 seconds and passed all 38,615 tests in 85 seconds, with 0 audit findings. ESLint remains the safest default for teams that value rule choice, plugins, and custom policy over a single preset.

Full review →
12

testify

stretchr/testify · ★ 26k1/7d · Go

Testify is a set of Go packages for assertions, stop-on-failure checks, mocks, and test suites. It keeps Go's standard `testing` runner, while cutting down the repetitive comparisons and failure messages developers otherwise write by hand.

SetupDocsCommunityMaturity

Verdict Our Testify run installed 3 packages in 6 seconds and passed all 8 tests, so it is an easy recommendation for Go teams that want clearer assertions without changing runners. Start with `assert` and `require`; add `mock` only at narrow interfaces.

Full review →
13

simple-icons

simple-icons/simple-icons · ★ 26k50/7d · JavaScript

Simple Icons is a catalog of more than 3,400 SVG marks for popular brands, normalized for small interface use. Designers can download individual files, while developers can use the same icons through npm, Composer, or a CDN without collecting and cleaning each logo themselves.

SetupDocsCommunityMaturity

Verdict Our run of Simple Icons installed 593 packages and passed all 31,083 tests, so it is a dependable default for products that need many brand marks in one predictable format. Use named, versioned imports and verify each brand's current usage rules.

Full review →
14

markdown-it

markdown-it/markdown-it · ★ 22k · TypeScript

markdown-it turns Markdown text into HTML in Node.js or a browser. It follows CommonMark, exposes the parsed token stream, and lets applications add syntax or rendering behavior through plugins.

SetupDocsCommunityMaturity

Verdict Our markdown-it run installed 402 packages in 20 seconds, built in 13 seconds, and passed all 307 tests in 23 seconds, with 0 audit findings. That makes it an easy recommendation for applications needing CommonMark plus controlled extensions.

Full review →
15

golangci-lint

golangci/golangci-lint · ★ 19k14/7d · Go

golangci-lint runs a large catalog of Go linters through one command, configuration file, cache, and reporting pipeline. It gives a team one CI entry point for compiler checks, bug finders, style rules, security analysis, and formatters instead of wiring each tool separately.

SetupDocsCommunityMaturity

Verdict Our golangci-lint run installed 400 packages, built in 67 seconds, and passed all 142 tests, making it the safest default of this review batch for a Go team that wants several analyzers. Use the released binary rather than building through `go install`, pin it beside the configuration, and begin with a small enabled set.

Full review →
16

task

go-task/task · ★ 16k30/7d · Go

Task is a cross-platform command runner and build tool configured with YAML Taskfiles. It replaces repeated shell commands and many small Makefiles with named tasks, dependencies, variables, platform conditions, file fingerprints, and reusable local or remote includes.

SetupDocsCommunityMaturity

Verdict Our Task run built in 71 seconds and passed all 15 test packages in 20 seconds, making it the cleanest repository result in this batch. Use it when a team needs a portable, discoverable command layer and can live with YAML plus parallel dependency semantics.

Full review →
17

bottom

ClementTsang/bottom · ★ 14k17/7d · Rust

bottom, run as `btm`, is a graphical system and process monitor inside the terminal. It combines live CPU, memory, network, disk, temperature, battery, and process views on Linux, macOS, and Windows, with configurable layouts and process controls.

SetupDocsCommunityMaturity

Verdict Our bottom build completed in 164 seconds and all 323 tests passed in 62 seconds, making it the cleanest source checkout in this batch. Install it when you want live local graphs and practical process controls in one cross-platform terminal interface.

Full review →
18

ohmyzsh

ohmyzsh/ohmyzsh · ★ 190k107/7d · Shell

Oh My Zsh is a framework for organizing and extending a Zsh configuration. It gives terminal users a ready-made collection of plugins, aliases, themes, update tooling, and documented customization paths, reducing the work of building a useful shell setup from scratch.

SetupDocsCommunityMaturity

Verdict Our sandbox could not run the installer, build, or tests because Shell was unsupported and no Dockerfile was present. Oh My Zsh is still an easy recommendation for Zsh users who want a mature, documented starting point and will curate their plugin list.

Full review →
19

flutter

flutter/flutter · ★ 179k84/7d · Dart

Flutter is Google's open-source toolkit for building user interfaces that run on mobile, web, desktop, and embedded devices from a single Dart codebase. It solves the costly and time-consuming problem of developing and maintaining separate applications for each platform, like iOS and Android, by allowing developers to write code once and deploy it everywhere.

SetupDocsCommunityMaturity

Verdict Flutter is a stellar choice for building beautiful, performant, cross-platform applications from a single codebase. Its promise of controlling every pixel is both its greatest strength for branded experiences and a potential weakness for native purists.

Full review →
20

go

golang/go · ★ 139k798/7d · Go

Go is a compiled programming language and this repository contains its compiler, command-line tools, runtime, and standard library. The language is designed for readable code, fast builds, concurrency, and shipping standalone server or command-line programs.

SetupDocsCommunityMaturity

Verdict Our Go source setup stopped after 10 seconds because the 1.24 image could not obtain the tree's requested 1.28.0 toolchain, so `master` is the wrong starting point for someone who merely wants to use Go. Install the current binary release for application work; clone this repository only for compiler, runtime, or standard-library development.

Full review →
21

PowerToys

microsoft/PowerToys · ★ 139k180/7d · C

PowerToys is a Windows 10 and 11 utility suite from Microsoft that puts more than 30 small desktop tools under one settings app. It handles jobs such as arranging windows, remapping keys, launching apps, bulk-renaming files, extracting screen text, and keeping a PC awake without making users assemble a folder of separate utilities.

SetupDocsCommunityMaturity

Verdict Our PowerToys checkout installed its Node-based GitHub artifacts helper in 13 seconds and passed that helper's tests in 127 seconds, but we did not build the Windows suite. Windows 10 and 11 users should try the signed installer if several of its 30-plus utilities solve recurring annoyances; the app earns its place when FancyZones, Keyboard Manager, Text Extractor, and the launch tools all get used.

Full review →
22

free-for-dev

ripienaar/free-for-dev · ★ 138k516/7d · HTML

free-for-dev is a large, community-maintained directory of hosted services with free tiers for developers and infrastructure work. It saves you from searching dozens of vendor pricing pages when you need a database, CI service, monitoring tool, API, or other hosted building block for a small project.

SetupDocsCommunityMaturity

Verdict Use free-for-dev as the first pass when a project needs a hosted service on a small budget. Its scope rules and current contribution flow make it more useful than an unfiltered coupon list, but a one-line entry is not due diligence.

Full review →
23

godot

godotengine/godot · ★ 117k395/7d · C++

Godot is a free, open-source engine for making 2D and 3D games from one editor. It gives developers the rendering, scene, scripting, export, and supporting tools needed to build games without assembling an engine from separate parts.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 6 seconds, but pytest collected 0 tests and exited 5. Godot is an unusually credible default for independent 2D and 3D development because its editor, export targets, source access, and MIT license come without royalties.

Full review →
24

TypeScript

microsoft/TypeScript · ★ 111k86/7d · Go

TypeScript adds static types and editor tooling to JavaScript, then emits ordinary JavaScript for browsers, servers, and other runtimes. Version 7 moves the compiler and language server to a native Go implementation while keeping npm as the normal installation channel.

SetupDocsCommunityMaturity

Verdict Our TypeScript checkout installed 445 packages in 31 seconds, but build and tests both failed within 7 seconds because the Node image had no Go executable, so contributors need the new native toolchain even though users still install from npm. TypeScript 7 is the default choice for typed JavaScript applications, especially large builds that can use its parallel compiler.

Full review →
25

playwright

microsoft/playwright · ★ 96k354/7d · TypeScript

Playwright is a browser automation framework for testing websites and scripting Chromium, Firefox, and WebKit through one API. It solves the costly problem of maintaining separate browser drivers while adding a test runner, automatic waiting, traces, and isolated browser sessions.

SetupDocsCommunityMaturity

Verdict Our Playwright checkout installed 613 packages and built in 34 seconds total, but its repository test run ended with 202 failures and 14,800 tests unrun. Use the published package for serious cross-browser application testing: its isolation, locators, traces, and three-engine API are hard to match in one project.

Full review →
26

uv

astral-sh/uv · ★ 90k191/7d · Rust

uv is a Rust-built command-line tool for Python environments, dependencies, lockfiles, scripts, tools, Python installations, builds, and publishing. It reduces the number of separate utilities a Python developer needs while offering a familiar `uv pip` interface for common pip workflows.

SetupDocsCommunityMaturity

Verdict Our uv source run built in 4 seconds and pip-audit found 0 known vulnerabilities, but pytest stopped after 5 seconds with 3 setup errors before any test passed. Existing projects should migrate in a branch and compare lockfiles, install outputs, indexes, editable packages, and CI behavior before switching.

Full review →
27

caddy

caddyserver/caddy · ★ 76k169/7d · Go

Caddy is a web server and reverse proxy that obtains and renews TLS certificates automatically. It gives small sites a short Caddyfile, while larger systems can use native JSON, an admin API, and compiled Go modules.

SetupDocsCommunityMaturity

Verdict Our Caddy build succeeded in 114 seconds, but the 78-second test run ended with 21 package results passing and 1 failing, so contributors should reproduce the failing package before shipping a custom binary. For a small website or self-hosted reverse proxy, Caddy's automatic HTTPS and concise configuration make it one of the easiest serious options to operate.

Full review →
28

protobuf

protocolbuffers/protobuf · ★ 72k23/7d · C++

Protocol Buffers is a language-neutral way to define structured messages and encode them as compact binary data. You write a `.proto` schema, run the `protoc` compiler, and use generated types so services in different languages agree on field names, numbers, and data types.

SetupDocsCommunityMaturity

Verdict Protocol Buffers remains the safest default for typed service messages across several languages, especially with gRPC. Its wire format and tooling are mature, but the benefit depends on disciplined schema ownership and generated-code updates.

Full review →
29

Mole

tw93/Mole · ★ 67k750/7d · Shell

Mole is a macOS command-line toolkit for cleaning storage, uninstalling applications, inspecting disk use, refreshing caches and services, and watching system health. It replaces several narrow maintenance utilities with one free, open-source CLI, while an optional paid native Mac app provides a graphical interface.

SetupDocsCommunityMaturity

Verdict Our build completed in 19 seconds, and all 4 supplied Go tests passed in 10 seconds. Mole is unusually thoughtful for a deletion-capable CLI: dry runs, confirmations, whitelists, history, and conservative skipping make it easier to trust than a blind cleanup script.

Full review →
30

tldr

tldr-pages/tldr · ★ 64k92/7d · Markdown

tldr-pages is a community-written collection of short command-line help pages built around practical examples. It helps you remember a useful invocation quickly, while leaving exhaustive flags and edge cases to the command's manual or upstream documentation.

SetupDocsCommunityMaturity

Verdict Our tldr-pages checkout installed 95 packages in 14 seconds, but its 67-second test run failed because `black` was unavailable. That contributor snag does not weaken the everyday product: tldr remains the fastest sensible first stop when you need one common command example.

Full review →
31

json

nlohmann/json · ★ 51k · C++

nlohmann/json gives C++ programs a JSON value that feels much like a standard container. You can add one header or use a package manager, then parse, inspect, change, and serialize JSON without running a separate service.

SetupDocsCommunityMaturity

Verdict Our run installed 36 packages and built the selected tooling successfully, but pytest collected 0 tests and exited 5, so it does not validate nlohmann/json's C++ test suite. Use the library when readable C++ and painless integration outrank parser speed, especially in ordinary application code.

Full review →
32

brew

Homebrew/brew · ★ 50k183/7d · Ruby

Homebrew is a package manager that gives developers one command-line place to find, install, update, and troubleshoot software. It organizes command-line packages as formulae and desktop applications as casks, while `brew bundle` helps describe a machine's tool set.

SetupDocsCommunityMaturity

Verdict Our measured install completed in 23 seconds and the build finished in 7 seconds. Homebrew is an easy recommendation for developers who want its formula-and-cask ecosystem and accept a conventional command-line workflow.

Full review →
33

julia

JuliaLang/julia · ★ 49k26/7d · Julia

Julia is a programming language built for technical computing, combining an interactive, high-level style with an explicit focus on performance. It gives researchers, engineers, and numerical programmers one language for exploratory work and serious software instead of forcing an early split between a convenient scripting tool and a lower-level implementation.

SetupDocsCommunityMaturity

Verdict Our build completed in 211 seconds, but the test command discovered 0 tests. Julia is a strong choice when technical computing is the center of the product and the team is ready to commit to its ecosystem, not merely sample its syntax.

Full review →
34

moment

moment/moment · ★ 48k · JavaScript

Moment.js is an English-documented JavaScript library for parsing, validating, changing, and formatting dates. It gave older browser and Node.js applications one familiar API for date math and localization, but its maintainers now call it a legacy project in maintenance mode.

SetupDocsCommunityMaturity

Verdict Our Moment checkout installed 314 packages and used 378 MB, then passed its build and tests in 28 seconds combined, so existing maintainers still have a working codebase. Keep it when a mature application already relies on Moment's behavior, and update promptly for v2.31.0's security fix.

Full review →
35

monaco-editor

microsoft/monaco-editor · ★ 47k48/7d · JavaScript

Monaco Editor is the browser code editor generated from the same source used by Visual Studio Code. It gives web applications a capable editing surface with models, language services, completion providers, hover providers, and a versioned TypeScript API.

SetupDocsCommunityMaturity

Verdict Our Monaco Editor checkout installed 698 packages, built in 157 seconds, and passed its tests in 90 seconds, making it the safest choice here for a desktop-class browser editor. Use it when VS Code-like editing behavior matters enough to justify worker setup and a 515 MB development install.

Full review →
36

cli

cli/cli · ★ 46k75/7d · Go

GitHub CLI, usually invoked as `gh`, puts GitHub pull requests, issues, releases, Codespaces, Actions, and repository operations in the terminal. It is the official command-line client for GitHub.com, GitHub Enterprise Cloud, and GitHub Enterprise Server 2.20 or newer.

SetupDocsCommunityMaturity

Verdict Our `cli/cli` build finished in 66 seconds, but its full Go test run ended after 382 seconds with 224 of 226 packages passing, so contributors should budget for a large suite and investigate the two unexplained failures. For daily GitHub work, `gh` is still the default recommendation because it covers the service broadly, ships signed immutable releases, and has current maintenance.

Full review →
37

bubbletea

charmbracelet/bubbletea · ★ 45k115/7d · Go

Bubble Tea is a Go framework for building interactive terminal applications with the Elm-style model, update, and view loop. It handles terminal input and screen rendering while your program keeps application state and decides how each message changes it.

SetupDocsCommunityMaturity

Verdict Our Bubble Tea run installed 21 packages in 28 seconds, built in 20 seconds, and passed all 2 available tests in 10 seconds. Use it for a Go terminal application when explicit state and message handling sound cleaner than a mutable widget tree.

Full review →
38

zod

colinhacks/zod · ★ 44k41/7d · TypeScript

Zod is a TypeScript-first library for describing the shape of data, checking unknown input, and deriving matching TypeScript types from the same definition. It solves the common problem of static types disappearing at runtime, especially at API, form, configuration, and storage boundaries.

SetupDocsCommunityMaturity

Verdict Our build finished in 17 seconds, and all 7,808 tests passed, a concrete sign that Zod's current codebase is in strong working order. Use it when TypeScript applications need trustworthy runtime boundaries without maintaining separate types and validators.

Full review →
39

vim

vim/vim · ★ 41k19/7d · Vim Script

Vim is a keyboard-driven editor for code and plain text, built as a much expanded version of the Unix vi editor. It solves the problem of editing quickly in terminals and on many desktop systems while retaining vi-compatible behavior for people and scripts that depend on it.

SetupDocsCommunityMaturity

Verdict Our Vim build finished in 48 seconds, but the generic test step exited with `no test target`, so packaged users get a safer path than first-time source contributors. Use Vim if vi motions, terminal reach, and long-lived configuration compatibility already matter to your work.

Full review →
40

croc

schollz/croc · ★ 40k62/7d · Go

croc is a command-line tool for sending files and folders directly between computers with end-to-end encryption. It replaces account-based cloud sharing and awkward network setup with a short code phrase, while a relay fallback helps peers connect when a direct path is unavailable.

SetupDocsCommunityMaturity

Verdict Our build completed in 130 seconds, and all 56 measured tests passed. croc is an unusually practical choice for one-off encrypted transfers because it combines short-code pairing, resume support, multiple files, browser interoperability, and relay fallback without demanding an account.

Full review →
41

esbuild

evanw/esbuild · ★ 40k18/7d · Go

esbuild is a JavaScript, TypeScript, JSX, and CSS bundler and minifier written in Go. It turns source modules into browser or Node.js output through a command line, JavaScript API, or Go API, with short rebuild cycles as its defining goal.

SetupDocsCommunityMaturity

Verdict Our esbuild checkout installed in 9 seconds, built in 31 seconds, and passed all 16 Go packages in 13 seconds. It is the low-level default I would try first for ordinary web bundling or fast transforms.

Full review →
42

zoxide

ajeetdsouza/zoxide · ★ 39k208/7d · Rust

Zoxide is a command-line directory jumper that learns which folders you use often and lets you return to them with a few characters. It replaces repeated long `cd` paths with ranked matches while keeping ordinary relative and absolute paths available.

SetupDocsCommunityMaturity

Verdict Our zoxide run installed 59 packages in 15 seconds, built in 26 seconds, and passed all 32 tests, making it an easy recommendation for people who live in a terminal. Its learned ranking is faster than typing long paths, but it can be awkward around similar parent and worktree paths.

Full review →
43

compose

docker/compose · ★ 38k40/7d · Go

Docker Compose turns a YAML description of several containers into a repeatable local application environment. It solves the everyday problem of starting services, networks, ports, and volumes together instead of managing each container by hand.

SetupDocsCommunityMaturity

Verdict Our build finished in 74 seconds and all 34 measured tests passed, a concrete sign that the checked-out commit is healthy. Docker Compose remains the sensible default for a multi-container development environment when the team already uses Docker.

Full review →
44

marked

markedjs/marked · ★ 37k · JavaScript

Marked turns Markdown text into HTML for use in Node.js, browsers, or a command-line workflow. It solves the common need to render Markdown quickly while leaving security filtering and page presentation to the rest of your application.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 14 seconds, and the full test command completed successfully in 31 seconds. Marked is an easy recommendation when you want a mature, focused Markdown-to-HTML component and are prepared to sanitize untrusted output yourself.

Full review →
45

awesome-docker

veggiemonk/awesome-docker · ★ 37k

Awesome Docker is a volunteer-maintained directory of open-source projects built specifically around Docker. It helps developers and operators find tools for building images, running containers, security, storage, observability, interfaces, and day-to-day workflow without starting from a generic web search.

SetupDocsCommunityMaturity

Verdict Our sandbox build succeeded in 33 seconds, and all 16 tests passed. Awesome Docker is one of the better starting maps for the Docker tool ecosystem because its inclusion rule is unusually clear and its categories follow real operational jobs.

Full review →
46

pnpm

pnpm/pnpm · ★ 37k61/7d · Rust

pnpm is a package manager for JavaScript and Node.js projects that stores package files once and links them into each project. It reduces duplicated disk use, enforces declared dependencies more strictly than a flat `node_modules`, and includes serious workspace and supply-chain controls.

SetupDocsCommunityMaturity

Verdict Our pnpm repository run installed 1 detected package in 96 seconds, but exposed no build or test target, so it tells us little about the 787,801-line monorepo's correctness. Even with that evidence gap, pnpm is our first choice for a new Node workspace because strict dependency visibility and shared storage solve daily problems.

Full review →
47

fish-shell

fish-shell/fish-shell · ★ 34k47/7d · Rust

Fish is an interactive command-line shell for macOS, Linux, BSD-family systems, and Windows through WSL, Cygwin, or MSYS2. It provides syntax coloring, suggestions from history, and command completions out of the box, reducing the configuration normally needed to make Bash or Zsh comfortable for daily terminal work.

SetupDocsCommunityMaturity

Verdict Our Fish harness installed 35 packages and built in 2 seconds, but Pytest collected 0 tests and exited 5, so this run does not validate the shell's documented Rust build or behavior. Fish is still an excellent daily shell for someone who values suggestions and completions more than POSIX syntax compatibility.

Full review →
48

WSL

microsoft/WSL · ★ 34k54/7d · C++

Windows Subsystem for Linux lets Windows users run Linux command-line tools and applications alongside their Windows software. This repository contains the open-source WSL application and its developer build loop, while the Linux kernel, GUI integration, and user documentation live in related repositories.

SetupDocsCommunityMaturity

Verdict Our run installed 50 Python packages and built WSL's `distributions/` helper in 6 seconds, but pytest collected 0 tests, so it does not prove that WSL itself builds or passes on Windows. For most Windows developers, WSL remains the first option to try because `wsl --install`, Microsoft ownership, active stable releases, and direct host integration are hard to match.

Full review →
49

CheatSheetSeries

OWASP/CheatSheetSeries · ★ 33k · Python

OWASP Cheat Sheet Series is a practical library of security guidance for people building and reviewing software. It turns broad security topics into focused references that teams can consult while designing features, writing code, and checking controls.

SetupDocsCommunityMaturity

Verdict Our sandbox installed 241 packages in 9 seconds and completed the repository tests in 18 seconds. Use OWASP Cheat Sheet Series as a trusted starting point and team reference, especially when you need actionable defensive guidance without buying a platform.

Full review →
50

delta

dandavison/delta · ★ 32k77/7d · Rust

Delta is a terminal pager that makes Git diffs, unified diffs, grep output, and blame output easier to read with syntax colors, word-level change marking, line numbers, and optional side-by-side layout. It solves a daily ergonomics problem without changing how Git calculates the diff.

SetupDocsCommunityMaturity

Verdict Our Delta run passed all 876 tests after a 29-second install and 73-second build, which is strong evidence for a terminal tool that sits in every Git review. Use it if you want line-oriented diffs to stay familiar while becoming much easier to scan.

Full review →
51

minikube

kubernetes/minikube · ★ 32k20/7d · Go

Minikube creates a local Kubernetes cluster on macOS, Linux, or Windows for developing and testing applications before they reach a shared cluster. It wraps virtual-machine and container drivers, Kubernetes versions, networking, storage, dashboards, and optional addons behind one command-line tool.

SetupDocsCommunityMaturity

Verdict Our Minikube run installed 529 packages and built in 160 seconds, but 10 of 82 test targets failed, with the tail explicitly asking for a missing latest binary. Use Minikube when you want the broadest local Kubernetes experience across drivers, addons, GPUs, and operating systems.

Full review →
52

turborepo

vercel/turborepo · ★ 31k19/7d · Rust

Turborepo is a task runner and cache for JavaScript and TypeScript repositories, especially monorepos with many apps and packages. It reads package relationships and a `turbo.json` task graph so builds, tests, lint jobs, and type checks run in the right order without repeating unchanged work.

SetupDocsCommunityMaturity

Verdict Our Turborepo install took 42 seconds and 1,896 packages, but the repository offered our lab no generic build or test target, so that run proves dependency setup only. Use Turborepo when repeated JavaScript workspace tasks are already wasting developer or CI time and someone will own the cache contract.

Full review →
53

spdlog

gabime/spdlog · ★ 30k20/7d · C++

spdlog is a C++ logging library for writing formatted messages to consoles, files, system logs, debuggers, Qt widgets, or custom destinations. It solves the routine work of log levels, formatting, rotation, thread safety, and optional background delivery without forcing an application to adopt a larger observability system.

SetupDocsCommunityMaturity

Verdict Our spdlog checkout built in 154 seconds and all 2 ctest targets passed in 22 seconds, making it the cleanest source run in this group. Use it as the default shortlist choice for ordinary C++ application logging, especially when console, rotating files, and custom sinks cover the job.

Full review →
54

claude-hud

jarrodwatts/claude-hud · ★ 28k114/7d · JavaScript

Claude HUD is a Claude Code plugin that puts context use, subscription limits, repository state, active tools, agents, and todo progress below the prompt. It helps developers see when a session is filling up and what Claude Code is doing without opening another terminal pane.

SetupDocsCommunityMaturity

Verdict Our Claude HUD run installed 75 packages in 5 seconds, passed all 1,073 tests, and found 0 known vulnerabilities, which makes it an easy trial for a regular Claude Code user. Keep the default view small, and enable activity rows only when they answer a question you often have.

Full review →
55

zstd

facebook/zstd · ★ 28k82/7d · C

Zstandard is a lossless compression format plus its reference C library and command-line program. It gives developers a tunable alternative to older general-purpose compressors, with a standardized format, fast decompression, and dictionary training for families of small records.

SetupDocsCommunityMaturity

Verdict Our zstd checkout installed in 16 seconds, built in 44 seconds, and passed its tests in 10 seconds, making it the easiest recommendation in this group for developers who need lossless compression. Use the library when you control both ends or your ecosystem already supports RFC 8878, and use the CLI when you want a capable local compressor.

Full review →
56

fmt

fmtlib/fmt · ★ 26k100/7d · C++

{fmt} is a C++ library for turning values into strings and printing them without the awkwardness and safety problems of printf or iostreams. It gives developers a consistent formatting syntax, compile-time checks where supported, and helpers for dates, containers, colors, files, Unicode, and custom types.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 109 seconds, and all 22 measured tests passed in 7 seconds. {fmt} is an easy recommendation for C++ teams that need consistent formatting across toolchains, especially when compile-time checking, custom types, or older compiler support matters.

Full review →
57

oh-my-posh

JanDeDobbeleer/oh-my-posh · ★ 23k31/7d · Go

Oh My Posh is a cross-platform prompt engine that gives many shells the same configurable status line. It turns the current path, Git state, language versions, cloud context, command result, and other data into themed blocks and segments, solving the problem of maintaining separate prompt scripts on Windows, macOS, and Linux.

SetupDocsCommunityMaturity

Verdict Our sandbox installed 82 packages in 35 seconds, built Oh My Posh in 34 seconds, and passed all 54 tests in 25 seconds, a convincing source-level result for a prompt engine. Use it when one expressive configuration across several shells is worth the font and profile setup.

Full review →
58

sops

getsops/sops · ★ 23k61/7d · Go

SOPS is a command-line editor for encrypted YAML, JSON, ENV, INI, and binary files. It lets teams keep encrypted secrets in Git while granting decryption through age or PGP keys, cloud key-management services, or Vault-compatible services.

SetupDocsCommunityMaturity

Verdict Our SOPS build passed in 74 seconds, but tests ended with 18 packages passing and 4 failing, including a PGP key-source case, so contributors should reproduce that path before trusting a source change. For users, SOPS remains an excellent fit when encrypted files must live in Git and access can be tied to age or cloud KMS identities.

Full review →
59

ratatui

ratatui/ratatui · ★ 23k87/7d · Rust

Ratatui is a Rust library for building interactive interfaces inside a terminal. It supplies layout, text, styling, widgets, buffers, and several terminal backends, leaving application state and event handling in your code.

SetupDocsCommunityMaturity

Verdict Our Ratatui run installed 445 packages, built cleanly, and passed all 6,013 tests. It is our default Rust TUI recommendation for developers who want precise control and are comfortable writing the application loop.

Full review →
60

PHPMailer

PHPMailer/PHPMailer · ★ 22k1/7d · PHP

PHPMailer is a PHP library for composing and sending email without hand-building fragile message headers and MIME bodies. It handles SMTP delivery, authentication, attachments, HTML alternatives, international text, and signing so application developers can focus on the message rather than email protocol details.

SetupDocsCommunityMaturity

Verdict Our 2026-09-10 sandbox run could not install, build, or test PHPMailer because the runner had no supported PHP ecosystem and the repository had no Dockerfile. Even without an executable lab result, the documented feature set, recent push, modest issue count, and long adoption history make PHPMailer a strong default for direct email work in plain PHP.

Full review →
61

chezmoi

twpayne/chezmoi · ★ 22k76/7d · Go

chezmoi is a command-line dotfile manager that keeps one desired configuration across Linux, macOS, Windows, BSD, and other machines. It stores a source copy in Git, renders machine-specific files with templates, previews differences, and applies the minimum changes needed to the home directory.

SetupDocsCommunityMaturity

Verdict Our detected chezmoi workflow installed 34 packages in 19 seconds, built in 1 second, and reported 0 known vulnerabilities, but it exposed no test target. chezmoi is the best fit here for experienced command-line users whose dotfiles vary by machine and may pull secrets at apply time.

Full review →
62

book

rust-lang/book · ★ 18k19/7d · Rust

The Rust Programming Language, usually called the Rust Book, is the official free guide for learning Rust. It explains the language through an ordered course that starts with Cargo and basic syntax, then works through ownership, error handling, traits, concurrency, async code, and larger programs.

SetupDocsCommunityMaturity

Verdict Our run installed 34 packages in 32 seconds, built in 38 seconds, and passed all 74 tests in 31 seconds, so working with the official Rust Book repository carries little setup risk. Start here if you learn well from a sustained technical text, then add Rustlings when ownership or lifetimes need hands-on repetition.

Full review →
63

rust-analyzer

rust-lang/rust-analyzer · ★ 17k27/7d · Rust

rust-analyzer is the language server that gives Rust editors completion, navigation, refactoring, formatting, and diagnostics. It separates Rust code intelligence from the editor, so VS Code, Vim, Emacs, Zed, and other LSP clients can share the same analysis engine.

SetupDocsCommunityMaturity

Verdict Our rust-analyzer run passed 2,275 of 2,277 tests, but both formatting-document cases returned null after a 160-second build. It remains the sensible default for Rust code intelligence in an LSP editor because it is current, editor-neutral, and released weekly.

Full review →
64

yq

mikefarah/yq · ★ 16k18/7d · Go

yq is a command-line tool for reading, selecting, changing, merging, and converting YAML and other structured text formats. It brings a jq-like expression language to YAML, JSON, XML, CSV, TOML, HCL, properties, and related files, including in-place edits that try to retain YAML comments and style.

SetupDocsCommunityMaturity

Verdict Our yq build finished in 31 seconds, but 1 of 4 Go test packages failed because an unknown extension defaulted to YAML instead of returning `unknown`, so pin the version and test format detection in automation. yq is still the first tool to try for readable YAML queries, merges, conversions, and in-place changes across platforms.

Full review →
65

cargo

rust-lang/cargo · ★ 15k20/7d · Rust

Cargo is Rust's standard package manager and build front end. It resolves and downloads crate dependencies, calls the Rust compiler with the right arguments, and gives projects shared commands for building, testing, documenting, and publishing.

SetupDocsCommunityMaturity

Verdict Our Cargo run installed 445 packages and built successfully, but its tests stopped after 143 seconds because the expected Cargo executable was missing from `PATH`. Rust application teams should still use Cargo by default; the failed run is a warning for people building Cargo itself, not evidence against ordinary crate development.

Full review →
66

rust-clippy

rust-lang/rust-clippy · ★ 14k · Rust

Clippy checks Rust code for common mistakes, suspicious constructs, slow patterns, and needlessly complex code. It ships through rustup and runs as a Cargo subcommand, giving most Rust projects access to more than 800 documented lints without a separate service.

SetupDocsCommunityMaturity

Verdict Our Clippy build succeeded in 108 seconds and all 38 tests passed in 297 seconds, making it the sensible default linter for a Rust team that accepts some policy work. Use its default groups in local development and CI, then add pedantic or restriction lints one at a time.

Full review →
67

fresh

sinelaw/fresh · ★ 8.9k140/7d · Rust

Fresh is a terminal text editor with familiar desktop-style controls instead of a required modal command language. It combines ordinary editing, language-server features, an integrated terminal, plugins, and git-worktree task switching in one Rust application.

SetupDocsCommunityMaturity

Verdict Our Fresh build completed in 16 seconds with 0 known dependency vulnerabilities, but its repository offered no test target to our harness. It is the terminal editor I would show first to a VS Code or Sublime Text user who now works over SSH.

Full review →
68

build-your-own-x

codecrafters-io/build-your-own-x · ★ 547k1.2k/7d · Markdown

Build Your Own X is a large, English-language index of tutorials for recreating familiar software, including databases, shells, browsers, neural networks, and operating systems. It helps developers find project-based explanations of how foundational tools work instead of learning only through API use.

SetupDocsCommunityMaturity

Verdict Build Your Own X is one of the best places to choose an ambitious weekend or month-long learning project, provided you are willing to judge the linked tutorial yourself. It is a catalog, not a curriculum, and the absence of difficulty labels makes browsing harder than it should be.

Full review →
69

awesome-python

vinta/awesome-python · ★ 321k1.5k/7d · Python

Awesome Python is an edited shortlist of Python frameworks, libraries, and developer tools, organized by the job each project does. Its website adds search and filters, making it useful for finding a few credible candidates before doing a deeper comparison.

SetupDocsCommunityMaturity

Verdict Our Awesome Python run installed 35 packages in 21 seconds and built the site in 4 seconds, but its tests hit 3 import errors before any case ran. The list itself remains an excellent first stop because it now limits each use case to a small editorial set instead of rewarding endless submissions.

Full review →
70

yt-dlp

yt-dlp/yt-dlp · ★ 191k1.4k/7d · Python

yt-dlp is a command-line program and Python library for downloading audio, video, subtitles, thumbnails, and metadata from thousands of sites. It handles site-specific extraction, format choice, playlists, authentication, and post-processing that a plain media URL downloader does not.

SetupDocsCommunityMaturity

Verdict Our yt-dlp run installed 36 packages in 27 seconds and built in 8 seconds, but pytest stopped after 200 failures, many tied to live sites that had changed or rejected requests. It remains the first tool to try for lawful, scriptable media downloads because its extractor coverage and format controls are hard to match.

Full review →
71

spec-kit

github/spec-kit · ★ 137k2.8k/7d · Python

Spec Kit is a command-line toolkit that gives AI coding agents a specification-first development process. It turns an idea into project principles, requirements, a technical plan, tasks, implementation work, and a convergence check so the written intent stays visible while code changes.

SetupDocsCommunityMaturity

Verdict Spec Kit is worth adopting for consequential agent-built features where the cost of a wrong interpretation exceeds the cost of writing and reviewing several artifacts. Its templates give a team common checkpoints, and the measured test result supports confidence in the CLI.

Full review →
72

react-native

react/react-native · ★ 127k87/7d · C++

React Native is a framework for building Android and iOS applications with React and JavaScript while rendering native platform controls. It lets product teams share much of their UI and application logic while still calling Swift, Objective-C, Java, Kotlin, or C++ when the shared layer is not enough.

SetupDocsCommunityMaturity

Verdict Our React Native checkout installed 1,045 packages, built in 14 seconds, and passed 5,845 of 5,846 Jest tests with 1 skipped. That is strong repository evidence for a large project, and React teams should shortlist it when they can retain native mobile expertise.

Full review →
73

node

nodejs/node · ★ 122k773/7d · JavaScript

Node.js is a cross-platform JavaScript runtime for servers, command-line programs, build tools, and network services. It lets developers run JavaScript outside a browser and supplies built-in APIs for files, processes, networking, cryptography, tests, and other system work.

SetupDocsCommunityMaturity

Verdict Node.js remains the safest default for a JavaScript backend when compatibility, hiring, library choice, and long support windows outweigh the appeal of a newer all-in-one runtime. Use an even-numbered LTS line for production and pin it deliberately.

Full review →
74

pi

earendil-works/pi · ★ 106k2.2k/7d · TypeScript

Pi is a terminal coding agent and a set of TypeScript packages for building agent loops, model integrations, and terminal interfaces. It gives developers a small default toolset and expects them to shape the workflow through extensions, skills, prompts, themes, or embedded APIs.

SetupDocsCommunityMaturity

Verdict Our Pi run installed 352 packages and built in 31 seconds, but its 322-second test step failed with 1 of 87 Vitest tests failing in the SQLite session backend. Pi is a strong choice for developers who want a moldable, multi-provider terminal harness and will supply their own sandbox and workflow rules.

Full review →
75

neovim

neovim/neovim · ★ 102k123/7d · Vim Script

Neovim is a terminal text editor that refactors Vim around Lua, asynchronous jobs, an embedded terminal, and APIs that outside programs can use. It keeps modal editing and broad Vim plugin compatibility while making the editor easier to extend with modern interfaces and language tooling.

SetupDocsCommunityMaturity

Verdict Our Neovim source install stopped after 1 second because CMake could not find a Lua 5.1-compatible interpreter, so use an official package unless you are prepared to check the build prerequisites. Neovim is still the best fit for Vim users who want Lua configuration, external interfaces, and a busy plugin ecosystem without leaving modal editing.

Full review →
76

nvm

nvm-sh/nvm · ★ 95k35/7d · Shell

nvm is a per-user shell function for installing and switching between Node.js versions. It solves the common problem of projects needing different Node releases without replacing the system installation or using `sudo` for global packages.

SetupDocsCommunityMaturity

Verdict Our nvm test run finished with 203 passing and 17 failing tests after 221 seconds, so the current checkout did not earn a clean source-level pass in a fresh Debian container. For interactive POSIX-shell development, nvm remains an easy default because its `.nvmrc` workflow and troubleshooting coverage are hard to beat.

Full review →
77

PayloadsAllTheThings

swisskyrepo/PayloadsAllTheThings · ★ 81k128/7d · Python

PayloadsAllTheThings is an English-language reference of web security attack techniques, payload examples, bypasses, tools, and practice links. It helps authorized testers move from identifying a vulnerability class to checking specific behaviors without rebuilding every test string from memory.

SetupDocsCommunityMaturity

Verdict PayloadsAllTheThings is an excellent bench reference for authorized web testing because it pairs concrete payloads with context, tools, labs, and references. Keep it beside a proxy, not inside an unattended production workflow.

Full review →
78

cpython

python/cpython · ★ 77k727/7d · Python

CPython is the reference implementation of the Python programming language, including the interpreter, standard library, C API, tests, and official documentation sources. Use this repository when you need to change Python itself, embed it, study its behavior, or contribute to the language rather than merely run Python applications.

SetupDocsCommunityMaturity

Verdict Our CPython documentation setup installed 56 packages in 31 seconds and built in 7 seconds, but it did not compile or test the interpreter. That makes the checkout easy to recommend for documentation work and essential for changes to Python itself, while offering no reason for an application developer to build from main.

Full review →
79

imgui

ocornut/imgui · ★ 76k66/7d · C++

Dear ImGui is a C++ immediate-mode GUI library that draws tool and debug interfaces by emitting vertex buffers into your existing 3D pipeline. It solves the pain of building internal editors, inspectors, and live-tuning widgets without a separate retained UI framework.

SetupDocsCommunityMaturity

Verdict We could not build Dear ImGui in our sandbox (no Dockerfile, no C++ runner), so there is no measured build time to cite. Still, 76,111 GitHub stars, a last push one day before this review, and a v1.92.9b release 41 days earlier make this the default choice for C++ debug and tool UI..

Full review →
80

SecLists

danielmiessler/SecLists · ★ 74k155/7d · PHP

SecLists is a large, ready-made collection of usernames, passwords, URLs, sensitive-data patterns, fuzzing payloads, web shells, and other lists used in authorized security assessments. It saves testers from rebuilding the same input dictionaries whenever they prepare a new testing box.

SetupDocsCommunityMaturity

Verdict We could not run SecLists in our 3-CPU, 8 GB Debian sandbox because the harness found no supported PHP ecosystem and the repository has no Dockerfile. Use it if you already know which lists your authorized workflow needs: its breadth, active maintenance, and easy package options make it an excellent shared baseline.

Full review →
81

starship

starship/starship · ★ 60k72/7d · Rust

Starship is a configurable command prompt that works across many shells and operating systems. It replaces separate Bash, Zsh, Fish, PowerShell, and other prompt setups with one executable and one TOML configuration that can show the current directory, Git state, language versions, cloud context, and command status.

SetupDocsCommunityMaturity

Verdict Our Starship run built successfully, but 96 of 1,254 executed tests failed, so the released binary is easier to recommend than this commit as a clean contributor baseline. Use it if one prompt must follow you across shells and you are prepared to tune or disable slow modules.

Full review →
82

awesome-rust

rust-unofficial/awesome-rust · ★ 59k109/7d · Rust

Awesome Rust is a curated directory of Rust applications, libraries, development tools, registries, and learning resources. It solves the early research problem: finding credible projects by topic without searching GitHub and crates.io from scratch.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 72 seconds, but cargo test discovered 0 tests. Use Awesome Rust as a high-quality discovery index, not as evidence that any listed dependency is safe or best in class.

Full review →
83

typst

typst/typst · ★ 56k102/7d · Rust

Typst is a markup-based typesetting system for producing polished documents without LaTeX's command-heavy syntax. Its open-source Rust compiler and CLI handle local compilation, live watching, fonts, math, bibliographies, and programmable layouts, while a separate hosted editor adds collaboration.

SetupDocsCommunityMaturity

Verdict Our Typst checkout passed all 42 tests, but its Rust build took 654 seconds, so the release binary is the sensible entry point for most authors. Use it for new technical documents when readable source, math, references, and programmable layout matter.

Full review →
84

first-contributions

firstcontributions/first-contributions · ★ 56k76/7d

First Contributions is a guided practice repository for people making their first open-source contribution. It turns the unfamiliar fork, branch, edit, commit, and pull-request sequence into a small, low-risk exercise, helping beginners learn the workflow before approaching a real project.

SetupDocsCommunityMaturity

Verdict Use First Contributions if the mechanics of a public pull request are still intimidating. Its narrow exercise, MIT license, enormous adoption, same-day repository activity, and unusually broad translation effort make it an easy recommendation for a first rehearsal.

Full review →
85

guava

google/guava · ★ 52k10/7d · Java

Guava is Google's collection of Java utilities for immutable collections, multimaps, multisets, graphs, hashing, concurrency, I/O, strings, and primitives. It fills gaps in the JDK with APIs that are widely reused in applications and libraries, with separate artifacts for regular JREs and Android.

SetupDocsCommunityMaturity

Verdict Our Guava test run passed 4,318,381 cases but still exited 1 with 3 failures, while the separate build failed on missing snapshot test-source artifacts. That contributor friction does not erase Guava's value as a stable Maven dependency: use the correct JRE or Android artifact and prefer non-beta APIs.

Full review →
86

tmux

tmux/tmux · ★ 49k115/7d · C

tmux is a terminal multiplexer that keeps shell sessions running after you disconnect and lets one terminal hold several windows and panes. It solves the familiar problem of losing remote work with an SSH connection, while giving keyboard-driven users a durable command-line workspace.

SetupDocsCommunityMaturity

Verdict Our tmux run installed in 17 seconds, built in 19 seconds, and passed its tests in 5 seconds, a clean result for a 132,255-line C project. Use it if terminal persistence and scriptable panes matter enough to learn its Ctrl-b command model.

Full review →
87

pandoc

jgm/pandoc · ★ 46k · Haskell

Pandoc is a command-line tool and Haskell library that converts documents from one markup format to another. It solves the recurring problem of turning a single source document into web pages, office files, ebooks, slides, and other deliverables without maintaining a separate authoring workflow for each format.

SetupDocsCommunityMaturity

Verdict We did not run Pandoc at commit c9a9a5e because our sandbox supports no Haskell ecosystem and the repository has no Dockerfile. That limits what we can claim about setup, but not what the supplied evidence says about its unusually broad conversion surface and current maintenance.

Full review →
88

i-have-adhd

ayghri/i-have-adhd · ★ 46k16k/7d · Python

i-have-adhd is an output-style skill for coding assistants that puts the next action first, numbers multi-step work, keeps progress visible, and cuts conversational filler. It does not diagnose or treat ADHD; it changes how an agent presents and carries work across a session.

SetupDocsCommunityMaturity

Verdict Our i-have-adhd install took 6 seconds and added 0 packages, but the repository exposed no lab build or test target, so adoption rests on how your model behaves with the rules active. Try it on demand before enabling its session-wide hooks.

Full review →
89

curl

curl/curl · ★ 43k47/7d · C

curl is a command-line program for moving data through URLs, and libcurl is the C library underneath it. Together they cover everyday HTTP requests, file transfers, mail protocols, proxies, authentication, TLS, and embedded network clients.

SetupDocsCommunityMaturity

Verdict Our curl source setup stopped after 8 seconds because Libpsl was missing, before any build or test could run. Install a distribution binary for ordinary command-line use; choose libcurl when its protocol range and stable C API justify deliberate dependency selection.

Full review →
90

black

psf/black · ★ 42k · Python

Black is an automatic Python formatter that rewrites whole files to one deliberately opinionated style. It removes recurring debates about spacing, wrapping, and quote style from code review, with only a small set of project options.

SetupDocsCommunityMaturity

Verdict Our Black install took 26 seconds and pip-audit found 0 known vulnerabilities, but the build exited 1 and the full suite ended with 445 of 449 tests passing plus 1 collection or setup error. Black is still the sensible default for a Python team that wants formatting settled by one command and can install the right development extras.

Full review →
91

trpc

trpc/trpc · ★ 41k21/7d · TypeScript

tRPC lets a TypeScript client call a TypeScript server with matching input, output, and error types, without maintaining a separate API schema or generated client. It is meant to catch broken API calls in the editor and at compile time while keeping the programming model close to ordinary functions.

SetupDocsCommunityMaturity

Verdict Our build finished in 31 seconds, and all 1,131 tests passed. Use tRPC when one TypeScript team controls both ends of an application and values fast feedback over a language-neutral contract.

Full review →
92

nushell

nushell/nushell · ★ 41k59/7d · Rust

Nushell is a cross-platform command shell that passes typed tables, records, lists, and other values through pipelines instead of treating everything as plain text. It lets people query files, processes, command output, and formats such as TOML with one expression language on Windows, macOS, and Linux.

SetupDocsCommunityMaturity

Verdict Our Nushell build succeeded, but it took 386 seconds and the suite ended with 60 failed tests after 359 seconds, despite 4,128 passes. Install the released binary if typed pipelines would replace piles of jq, awk, and platform-specific parsing in your daily work.

Full review →
93

ShareX

ShareX/ShareX · ★ 40k85/7d · C#

ShareX is a free, open-source Windows app for capturing screenshots, recording screens, annotating images, and sharing files. It replaces a pile of separate capture, editing, OCR, upload, and desktop utility tools, while letting power users automate what happens after every capture.

SetupDocsCommunityMaturity

Verdict Use ShareX if Windows screenshots are part of your actual work rather than an occasional keyboard shortcut. Its unusual strength is not any single capture mode, but the ability to turn capture, editing, saving, uploading, and link copying into a repeatable workflow.

Full review →
94

trivy

aquasecurity/trivy · ★ 38k82/7d · Go

Trivy is a command-line security scanner for container images, filesystems, Git repositories, virtual machine images, and Kubernetes. It finds known vulnerabilities, exposed secrets, infrastructure configuration mistakes, licenses, and software components for an SBOM, giving teams one tool for several build and deployment checks.

SetupDocsCommunityMaturity

Verdict Our Trivy build took 210 seconds and 213 of 216 test packages passed, but 3 failures keep commit dc3c56e from earning an unqualified source-build recommendation. The released tool remains a strong default for teams that want one CI scanner across images, repositories, secrets, SBOMs, and Kubernetes.

Full review →
95

CyberChef

gchq/CyberChef · ★ 36k69/7d · JavaScript

CyberChef is a browser workbench for decoding, encoding, decompressing, hashing, parsing, and inspecting data. You build a recipe by arranging operations, feed it text or a file, and see each transformation without writing a one-off script.

SetupDocsCommunityMaturity

Verdict Our CyberChef build and tests passed, but the 31-second install pulled 1,385 packages and npm audit reported 42 advisories, including 9 high-severity findings. Use it as a fast analyst's workbench, preferably from a pinned current release and with sensitive recipes kept out of shared URLs.

Full review →
96

zellij

zellij-org/zellij · ★ 35k75/7d · Rust

Zellij is an English-documented terminal workspace that keeps shells and programs in persistent tabs and panes, with layouts, shared sessions, WebAssembly plugins, and a built-in web client. It solves the problem of juggling many terminals while preserving a workspace you can detach from and rejoin.

SetupDocsCommunityMaturity

Verdict Our Zellij run resolved 436 packages but failed after 307 seconds because 12 WebAssembly plugin artifacts were missing, so source work should follow the documented `cargo xtask` path instead of a generic Cargo build. Most users should install the v0.45.0 binary and try its panes, nested sessions, and shell-aware scrollback.

Full review →
97

kitty

kovidgoyal/kitty · ★ 35k79/7d · Python

kitty is a cross-platform, GPU-based terminal emulator for people who spend much of their day at a command line. It aims to make terminal work fast and feature-rich without turning the terminal itself into a hosted service or a browser application.

SetupDocsCommunityMaturity

Verdict Our build completed in 7 seconds after a 16-second install, which makes kitty unusually easy to validate at the source-build level for a project of this size. You should shortlist it if you want an actively maintained, GPU-based terminal with a large user base and dedicated documentation.

Full review →
98

poetry

python-poetry/poetry · ★ 34k8/7d · Python

Poetry is a Python project manager that resolves and locks dependencies, creates isolated environments, builds packages, and publishes them. It puts work that often spans pip, virtualenv, build frontends, and several packaging files behind one command-line tool and a pyproject.toml file.

SetupDocsCommunityMaturity

Verdict Our Poetry checkout installed 70 packages in 67 seconds and built in 9 seconds, but its test command stopped after 6 seconds because `responses` was missing. Poetry remains a sensible default for Python teams that want dependency locking, environments, packaging, and publishing under one mature interface.

Full review →
99

mise

jdx/mise · ★ 34k290/7d · Rust

mise is a command-line tool that manages development-tool versions, project environment variables, and repeatable tasks from configuration committed with the code. It can replace separate version managers and task wrappers while giving local shells and CI the same declared setup.

SetupDocsCommunityMaturity

Verdict Our mise workspace install added 283 packages and 325 MB, but the 6-second test failure was only a placeholder script, so it provides no verdict on the Rust CLI's correctness. mise is an excellent fit for polyglot teams willing to make one tool responsible for runtimes, environment, and tasks, especially when migrating from asdf.

Full review →
100

comprehensive-rust

google/comprehensive-rust · ★ 33k10/7d · Rust

Comprehensive Rust is an open-source course from Google's Android team for experienced software engineers learning Rust. Its core curriculum runs for four days, with separate material on Android, Chromium, bare-metal work, concurrency, idiomatic code, and unsafe Rust.

SetupDocsCommunityMaturity

Verdict Our sandbox installed 321 packages in 24 seconds, built in 143 seconds, and passed all 107 tests in 11 seconds. This is an excellent ready-made curriculum for an experienced engineer teaching other experienced engineers, especially when Android or Chromium material is useful.

Full review →
101

atuin

atuinsh/atuin · ★ 32k63/7d · Rust

Atuin replaces flat shell history with a searchable SQLite database that records the command, time, working directory, exit status, duration, host, and session. It can stay entirely local or synchronize end-to-end encrypted history through Atuin Cloud or a server you operate.

SetupDocsCommunityMaturity

Verdict Our Atuin run passed 297 of 300 tests before three server tests hit connection resets, so commit 824c871 is close to clean but still needs that integration failure understood. The local search and metadata are useful even with sync disabled, which makes Atuin easier to recommend than a cloud-only history service.

Full review →
102

openai-python

openai/openai-python · ★ 32k23/7d · Python

OpenAI's official Python library gives Python 3.10+ applications a typed way to call the OpenAI REST API. It handles client setup, synchronous and asynchronous requests, authentication, and response objects so teams do not have to maintain their own HTTP wrapper.

SetupDocsCommunityMaturity

Verdict Our sandbox install succeeded in 15 seconds and occupied 37 MB. Use this SDK when a Python 3.10+ service talks directly to OpenAI and first-party types, auth options, and API coverage matter.

Full review →
103

Rectangle

rxhanson/Rectangle · ★ 30k51/7d · Swift

Rectangle is a free macOS window manager that moves and resizes windows with keyboard shortcuts or edge dragging. It solves the everyday problem of arranging apps into halves, thirds, quarters, and other practical layouts without manually adjusting each window.

SetupDocsCommunityMaturity

Verdict We did not run Rectangle at commit bf86a4b because our 3-CPU, 8 GB Debian sandbox had no supported Swift/macOS execution path and the repo had no Dockerfile. For most Mac users who want predictable snapping without adopting a full tiling system, Rectangle is an easy recommendation based on its focused feature set, recent v1.100 release, and active repository.

Full review →
104

composer

composer/composer · ★ 30k8/7d · PHP

Composer is the dependency manager for PHP projects. It reads a project’s declared packages, resolves compatible versions, and installs the resulting dependency set so teams do not have to manage third-party code by hand.

SetupDocsCommunityMaturity

Verdict We did not run commit 2616959 because our sandbox had no supported PHP path and the repository supplied no Dockerfile. Even without a lab result, Composer 2.10.3 is the sensible default for modern PHP dependency management: it is focused, actively maintained, and backed by the Packagist workflow.

Full review →
105

emscripten

emscripten-core/emscripten · ★ 28k7/7d · C++

Emscripten is a compiler toolchain that turns C and C++ programs into WebAssembly, with JavaScript or HTML needed to load them in browsers and other runtimes. It solves the hard parts of moving an existing native codebase to the web, including support for familiar libraries such as SDL2 and OpenGL-style graphics APIs.

SetupDocsCommunityMaturity

Verdict Our Emscripten checkout installed 258 npm packages in 14 seconds, but it exposed no npm build or test target, so a source clone is not a self-contained compiler setup. Use Emscripten when preserving a substantial C or C++ codebase is worth adapting its operating-system assumptions to browser rules.

Full review →
106

vcpkg

microsoft/vcpkg · ★ 27k11/7d · CMake

vcpkg is a free C and C++ package manager maintained by Microsoft and the wider community. It helps teams fetch, build, version, and reuse native dependencies across Windows, macOS, and Linux without maintaining a separate hand-written build recipe for every library.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 12 seconds. vcpkg is a strong default for cross-platform C and C++ teams that value a large port ecosystem, first-party CMake and MSBuild guidance, and reproducible dependency policy.

Full review →
107

angular-cli

angular/angular-cli · ★ 27k1/7d · TypeScript

Angular CLI creates, serves, tests, builds, and upgrades Angular applications from the command line. It gives Angular teams one supported workflow for workspace configuration, code generation, development servers, production bundles, and server-side rendering.

SetupDocsCommunityMaturity

Verdict Our Angular CLI install added 1,539 packages and used 909 MB, while both the 125-second build and 73-second test run failed on Bazel archive downloads. Use the published CLI for Angular work because its generators, builders, and migrations are the framework's supported path.

Full review →
108

openapi-generator

OpenAPITools/openapi-generator · ★ 27k7/7d · Java

OpenAPI Generator reads an OpenAPI description and produces client SDKs, server scaffolds, API documentation, or related configuration for many languages. It saves teams from hand-writing repetitive request models and endpoints, while leaving them responsible for checking the generated code.

SetupDocsCommunityMaturity

Verdict Our measured sample installed in 12 seconds and only used 6 MB, but its audit found 29 known vulnerabilities and it had no build or test target. OpenAPI Generator is still the practical shortlist choice for broad, multi-language generation, provided every output is compiled and tested as owned code.

Full review →
109

elixir

elixir-lang/elixir · ★ 27k11/7d · Elixir

Elixir is a programming language for applications that must handle many things happening at once and recover when part of the system fails. It runs on the Erlang virtual machine and includes its own build tool, test framework, interactive shell, and package workflow.

SetupDocsCommunityMaturity

Verdict Our lab did not run Elixir commit 1d15942 because the harness does not support the Elixir ecosystem and the repository supplies no Dockerfile. Elixir v1.20.4 still belongs on the shortlist for concurrent, fault-tolerant services when a team is prepared to operate the Erlang VM and learn the process model.

Full review →
110

difftastic

Wilfred/difftastic · ★ 26k35/7d · Rust

Difftastic compares code by syntax instead of treating every file as a stack of lines. It makes refactors and formatting changes easier to review because it can show which expressions changed even when they moved across lines.

SetupDocsCommunityMaturity

Verdict Our Difftastic build took 199 seconds and all 286 tests passed, a clean result for a developer tool best used beside Git's ordinary diff. Install the binary if reviews of reformatted or rearranged code cost your team time.

Full review →
111

biome

biomejs/biome · ★ 26k76/7d · Rust

Biome is a formatter and linter for JavaScript, TypeScript, JSX, JSON, CSS, and GraphQL, exposed through a CLI and language server. It aims to replace common Prettier and ESLint tasks with one Rust-based tool that can format, diagnose, and safely fix code from the editor or CI.

SetupDocsCommunityMaturity

Verdict Our Biome checkout installed 313 pnpm packages in 28 seconds and occupied 305 MB, but the root workspace offered no build or test target for the lab to verify. The published CLI is worth a controlled migration when one formatter-linter can replace most of a project's current Prettier and ESLint policy.

Full review →
112

core-js

zloirock/core-js · ★ 26k · JavaScript

core-js supplies JavaScript features that a target browser or runtime does not provide correctly yet. It lets applications use modern language and selected web-platform APIs while supporting older or inconsistent environments, either by patching globals or through imports that avoid global changes.

SetupDocsCommunityMaturity

Verdict Our test command failed after 313 seconds because codespell was not found. That contributor-experience snag does not erase core-js's central strength: it is a mature, modular compatibility layer with current development and unusually broad coverage.

Full review →
113

d2

d2lang/d2 · ★ 25k75/7d · Go

D2 is a text language and command-line renderer for making diagrams that can live beside source code and documentation. You describe objects and connections in a `.d2` file, then D2 arranges them and exports the result as an image, document, or presentation file.

SetupDocsCommunityMaturity

Verdict Our D2 run installed 60 packages in 31 seconds and built in 75 seconds, but 2 of 138 Go test packages failed. D2 is still an easy recommendation for teams that want attractive architecture diagrams in version control and can verify the rendered output in CI.

Full review →
114

lvgl

lvgl/lvgl · ★ 25k54/7d · C

LVGL is an MIT-licensed C library for building graphical interfaces on microcontrollers, embedded Linux devices, and other systems with a display. It supplies widgets, layouts, rendering, and input handling so firmware teams do not have to create a UI stack from raw pixels.

SetupDocsCommunityMaturity

Verdict Our LVGL checkout installed 44 Python packages in 23 seconds and built in 9 seconds, but pytest collected 0 tests and exited 5, so the measured developer path is incomplete. LVGL is still the first library I would trial for a C-based embedded display because its runtime is MIT licensed, hardware coverage is wide, and the manual port contract is explicit.

Full review →
115

Kingfisher

onevcat/Kingfisher · ★ 24k · Swift

Kingfisher is a Swift library that downloads, displays, processes, and caches remote images in Apple-platform apps. It replaces the repeated networking, cache, placeholder, transition, and image-view plumbing that teams would otherwise have to build themselves.

SetupDocsCommunityMaturity

Verdict Kingfisher 8.12.0 was not run in our sandbox because Swift was unsupported there and the repository had no Dockerfile. Even with that verification gap, its focused API, layered cache, image processors, SwiftUI path, and active maintenance make it an easy shortlist choice for modern Apple apps.

Full review →
116

HandBrake

HandBrake/HandBrake · ★ 24k77/7d · C

HandBrake converts existing video into MP4, MKV, or WebM files that are easier to store and play across phones, televisions, computers, and browsers. It wraps a complicated set of codecs and filters in desktop applications and a command-line tool, so most users can start with a preset instead of designing an encoding command.

SetupDocsCommunityMaturity

Verdict Our HandBrake source install stopped after 6 seconds because `libtool` was missing, so the repository is a poor first route for someone who only wants to convert a video. Install the official build and use HandBrake when presets, visible queue management, and broad playback compatibility matter.

Full review →
117

floci

floci-io/floci · ★ 24k562/7d · Java

Floci is a local AWS emulator for development, integration tests, and CI. It accepts familiar AWS SDK, CLI, Terraform, OpenTofu, and CDK requests on one local endpoint, then handles many services in-process or starts real Docker containers where a shallow mock would be misleading.

SetupDocsCommunityMaturity

Verdict Our Floci CDK compatibility build finished in 11 seconds, but the package had no test target and npm audit found 1 high-severity vulnerability. Floci is the most interesting choice for teams leaving LocalStack Community and needing broad, token-free local AWS behavior.

Full review →
118

gson

google/gson · ★ 24k · Java

Gson is an Apache-2.0 Java library that converts Java objects to JSON and back, including pre-existing classes and generic types. It solves routine data binding with reflection, a tree model, streaming readers and writers, and custom adapters when the defaults do not match your wire format.

SetupDocsCommunityMaturity

Verdict Our Gson run built in 75 seconds and recorded 16,892 passing tests, but the command still exited 1 when the JPMS integration module failed. Gson 2.14.0 remains a sensible choice for conventional Java 8 services that value compatibility and custom adapters over a growing feature set.

Full review →
119

coreutils

uutils/coreutils · ★ 24k36/7d · Rust

uutils coreutils rewrites the familiar GNU command-line utilities in Rust and targets Linux, macOS, BSD, Windows, and WASI. It gives operating systems and tool builders one portable codebase for commands such as `ls`, `cp`, and `sort`, while treating output or exit-code differences from GNU as bugs.

SetupDocsCommunityMaturity

Verdict Our uutils run built in 102 seconds and passed all 10,276 tests, which earns it a serious trial for portable systems work. Use it when Rust, a multicall binary, or Windows and WASI support solves a real packaging problem.

Full review →
120

grpc-go

grpc/grpc-go · ★ 23k9/7d · Go

gRPC-Go is the Go implementation of gRPC, an RPC framework that lets services call typed methods over HTTP/2. It generates client and server interfaces from protocol definitions so teams can share contracts across services and languages.

SetupDocsCommunityMaturity

Verdict Our grpc-go run built in 68 seconds and passed all 208 reported tests, making it the safest default for Go teams that have already chosen gRPC. Use it for typed internal APIs, streaming, and cross-language contracts where the supporting proxy and schema discipline are justified.

Full review →
121

witr

pranshuparmar/witr · ★ 22k195/7d · Go

witr is a command-line and terminal interface that traces a process, port, container, or open file back through the chain that started it. It combines process, service-manager, container, socket, and working-directory clues so an operator can answer why something is running without manually joining several system tools.

SetupDocsCommunityMaturity

Verdict Our witr run built in 22 seconds and passed all 7 tests in 11 seconds, so the small Go tool is cheap to evaluate. Install it if mystery ports and supervisor chains routinely send you through `ps`, `lsof`, service managers, and container CLIs.

Full review →
122

Inquirer.js

SBoudrias/Inquirer.js · ★ 22k · TypeScript

Inquirer.js is a set of TypeScript packages for asking questions inside interactive Node.js command-line programs. It supplies input, select, checkbox, confirm, search, password, number, editor, expand, and raw-list prompts, plus lower-level tools for custom prompts and automated terminal tests.

SetupDocsCommunityMaturity

Verdict Our Inquirer.js run passed 388 of 411 tests, but 22 failures and 13 snapshot mismatches mean source contributors should reproduce the suite before trusting this checkout. For application authors, the current `@inquirer/prompts` packages remain a strong default when a Node CLI genuinely needs a human at a terminal.

Full review →
123

CodexBar

steipete/CodexBar · ★ 21k343/7d · Swift

CodexBar is a macOS 14+ menu bar app that shows remaining quotas, reset times, credits, spend, and service status for AI coding providers. It combines many provider-specific APIs, browser sessions, command-line credentials, and local usage files into one small status view, with a separate CLI for macOS and Linux.

SetupDocsCommunityMaturity

Verdict CodexBar puts 69 provider integrations behind one macOS menu item, which is unusually useful and also a large maintenance bet on changing upstream interfaces. Use it when subscription limits interrupt real work and you can approve each provider's local credential source.

Full review →
124

mypy

python/mypy · ★ 21k · Python

Mypy is a static type checker that finds mismatched Python values and function calls before the program runs. It lets teams add type hints gradually, so an existing dynamic codebase can gain earlier feedback without being rewritten all at once.

SetupDocsCommunityMaturity

Verdict The build completed in 5 seconds, but the test suite did not finish within our 900-second cap. Mypy remains an easy recommendation for Python teams that want gradual, configurable static checking and are prepared to treat annotations as maintained code.

Full review →
125

grpc-gateway

grpc-ecosystem/grpc-gateway · ★ 20k · Go

gRPC-Gateway generates a Go proxy that turns ordinary JSON over HTTP requests into calls to an existing gRPC service. It lets one protobuf service support gRPC clients, REST-style clients, and generated API documentation without maintaining a second set of handlers by hand.

SetupDocsCommunityMaturity

Verdict Our gRPC-Gateway run installed 162 packages in 59 seconds, built in 55 seconds, and passed all 30 tests in 15 seconds, so the contributor path was clean on our box. Use it when protobuf already defines the service and ordinary HTTP clients need access; the mature Swagger 2.0 path is the safer default.

Full review →
126

magika

google/magika · ★ 19k88/7d · Rust

Magika identifies a file's content type from its bytes instead of trusting the filename or extension. It uses a small machine-learning model behind a Rust command-line tool and Python, Rust, JavaScript, and work-in-progress Go bindings, returning a stable label plus confidence and type metadata for routing files through other systems.

SetupDocsCommunityMaturity

Verdict Our Magika Go run installed 2 packages in 7 seconds, built in 12 seconds, and passed 2 of 2 tests in 14 seconds. The CLI or Python package is the safer default for most users because the Go binding still needs cgo, ONNX Runtime, linker flags, and separately packaged assets.

Full review →
127

wgpu

gfx-rs/wgpu · ★ 18k114/7d · Rust

wgpu is a Rust graphics library that gives native and browser applications one WebGPU-shaped API over Vulkan, Metal, Direct3D 12, OpenGL, WebGL2, and browser WebGPU. It handles the low-level differences between those graphics backends while keeping Rust's safety model at the public API boundary.

SetupDocsCommunityMaturity

Verdict Our wgpu run built in 198 seconds and passed 357 of 363 tests; all 6 reported failures said the external `spirv-val` executable was missing. The abstraction is broad, the project is active, and its testing architecture treats driver differences as real engineering work.

Full review →
128

capacitor

ionic-team/capacitor · ★ 17k58/7d · TypeScript

Capacitor turns an existing web app into iOS and Android projects and gives its JavaScript code a bridge to native device APIs. It helps web teams share one interface across mobile and the browser, while leaving platform-specific code and store builds in Xcode and Android Studio.

SetupDocsCommunityMaturity

Verdict Our Capacitor checkout installed 910 packages in 85 seconds and npm audit found 31 vulnerabilities, including 1 critical, so teams should treat it as a native platform dependency rather than a thin wrapper. Use it when a proven mobile web interface is the asset you want to preserve and the team can own Xcode and Android Studio projects.

Full review →
129

ast-grep

ast-grep/ast-grep · ★ 16k78/7d · Rust

ast-grep is a command-line search, lint, and code-rewriting tool that matches the structure of source code instead of raw text. You write a pattern that looks like normal code, add placeholders for the parts that may vary, and use the matches for one-off changes or repeatable project rules.

SetupDocsCommunityMaturity

Verdict Our ast-grep install took 271 seconds, but its build finished in 4 seconds and pip-audit found 0 known vulnerabilities. Use it for syntax-aware searches, repository rules, and codemods that can be expressed from tree structure alone.

Full review →
130

awesome-dsh-plugin

awesome-dsh-plugin/awesome-dsh-plugin · ★ 16k770/7d · Python

Awesome DSH Plugin is a bilingual English and Chinese directory of installable plugins for DeepSeek Harness. It helps DSH users find extensions for models, tools, memory, browser work, themes, security, workflows, and other parts of an agent system, while checking that each listing installs and matches its one-line description.

SetupDocsCommunityMaturity

Verdict Our Awesome DSH Plugin checkout needed only 3 packages and 3 MB, but it exposed no build or test target beyond that install. It is the best first directory for an existing DeepSeek Harness user because each listing must install and its description is checked against source.

Full review →
131

rolldown

rolldown/rolldown · ★ 14k5/7d · Rust

Rolldown is a Rust bundler for JavaScript and TypeScript, built to become Vite's production bundling engine. It keeps much of Rollup's API and plugin model while handling TypeScript, JSX, CommonJS, module resolution, code splitting, and minification itself.

SetupDocsCommunityMaturity

Verdict Our Rolldown run installed 1,103 packages in 72 seconds, then passed its 4-second build and 12-second test step, so trying the source is easier than its 83,812-file checkout suggests. Vite teams and slow Rollup builds should evaluate it, with production bundle execution and chunk graphs in the acceptance test.

Full review →
132

opa

open-policy-agent/opa · ★ 12k15/7d · Go

Open Policy Agent, or OPA, is an open-source engine that answers policy questions using rules written in Rego. Applications and infrastructure tools send it facts, ask whether an action or configuration is allowed, and then enforce the returned decision themselves.

SetupDocsCommunityMaturity

Verdict Our OPA run built in 110 seconds and passed 121 of 122 package test results, but the suite still exited 1, so commit 551581f is close to clean rather than verified clean. OPA remains a strong choice when one policy language truly needs to span several enforcement points and a team will own Rego and bundle operations.

Full review →
133

osv-scanner

google/osv-scanner · ★ 11k34/7d · Go

OSV-Scanner is a command-line vulnerability scanner that matches lockfiles, source trees, installed system packages, and container contents against the public OSV advisory database. It helps developers find known dependency problems and, for a smaller set of package formats, examine possible upgrades.

SetupDocsCommunityMaturity

Verdict Our OSV-Scanner build passed in 97 seconds and all 35 counted tests passed in 199 seconds after a 474-package install. It is an easy recommendation for multi-ecosystem dependency checks, especially when an open advisory database and offline mode matter.

Full review →
134

syft

anchore/syft · ★ 9.6k20/7d · Go

Syft inventories packages and files inside container images, archives, and directories, then writes a software bill of materials in formats such as SPDX, CycloneDX, or Syft JSON. It answers what software is present; a separate scanner such as Grype decides whether known vulnerabilities match that inventory.

SetupDocsCommunityMaturity

Verdict Our Syft build passed in 141 seconds, but 5 of 20 tests failed in 150 seconds and the final log showed a missing Docker executable during an image-based CLI test. Use Syft when you need broad, scriptable SBOM generation and will validate the output against the ecosystems and relationships your policy cares about.

Full review →
135

oapi-codegen

oapi-codegen/oapi-codegen · ★ 8.6k8/7d · Go

oapi-codegen turns an OpenAPI 3.0 or 3.1 document into Go types, API clients, and server interfaces. It removes repetitive request parsing and response wiring while leaving business logic in code your team owns.

SetupDocsCommunityMaturity

Verdict Our oapi-codegen run installed 56 packages and passed its build plus all 6 test targets in 40 seconds, making it the cleanest trial in this group. Use it for a Go codebase that owns its OpenAPI contract and will review committed generated diffs.

Full review →
136

trippy

fujiapple852/trippy · ★ 7.9k43/7d · Rust

Trippy is a terminal network diagnostic that combines repeated ping statistics with traceroute hop discovery. It traces with ICMP, UDP, or TCP, shows changing routes and latency in an interactive interface, and can export results for scripts or incident notes.

SetupDocsCommunityMaturity

Verdict Our Trippy build took 87 seconds and all 1,616 tests passed, making it the best-tested project in this group and an easy recommendation for interactive route diagnosis. Use it when protocol choice, ECMP visibility, and exportable reports justify raw-socket privileges.

Full review →
137

worktrunk

max-sixty/worktrunk · ★ 7.8k913/7d · Rust

Worktrunk is a Rust command-line tool that makes Git worktrees easier to create, inspect, merge, and remove. It is designed for developers running several coding agents or feature branches at once, with hooks, per-worktree status, shell integration, and agent activity markers.

SetupDocsCommunityMaturity

Verdict Our Worktrunk build took 169 seconds and 3,054 of 3,056 tests passed, with both failures leaving the cargo test command at exit 101. It is a good choice for experienced Git users who run parallel agents often enough to justify automated setup, status, and cleanup.

Full review →
138

pyrefly

facebook/pyrefly · ★ 7.0k20/7d · Rust

Pyrefly is a Python type checker and language server written mainly in Rust. It finds type mistakes from the command line and supplies editor features such as completion, navigation, hovers, and inlay hints, with migration tools for projects already using mypy or Pyright.

SetupDocsCommunityMaturity

Verdict Our Pyrefly build took 248 seconds and the test suite was still running at the 900-second cap, so adopters get a serious codebase with a contributor loop that needs patience. The packaged CLI is worth testing beside mypy or Pyright when one type engine across CI and the editor is attractive.

Full review →
139

cosign

sigstore/cosign · ★ 6.3k16/7d · Go

Cosign signs and verifies container images, blobs, and other OCI artifacts through the Sigstore ecosystem. It supports identity-based keyless signing, local encrypted keypairs, hardware or cloud key management, in-toto attestations, and signature storage alongside artifacts in a registry.

SetupDocsCommunityMaturity

Verdict Our Cosign run installed 788 packages, built in 150 seconds, and passed all 98 tests, so the codebase earns a serious production evaluation. Use v3.1.3 or newer, sign digests, and write verification policy around exact identities, issuers, keys, and attestation types.

Full review →
140

newton

newton-physics/newton · ★ 5.6k38/7d · Python

Newton is a physics simulation engine for robotics and simulation research, built on NVIDIA Warp with MuJoCo Warp as its main backend. It models robots, rigid and deformable bodies, cables, cloth, contacts, sensors, and parallel worlds through Python, with GPU acceleration on supported NVIDIA hardware.

SetupDocsCommunityMaturity

Verdict Our Newton install took 36 seconds and 520 MB, and its build passed in 5 seconds, but the harness found no test target to run. It deserves a serious trial for NVIDIA-backed robotics simulation, especially when batched worlds or Warp extensibility solve a measured problem.

Full review →
141

mcp-grafana

grafana/mcp-grafana · ★ 3.5k24/7d · Go

Grafana MCP server lets an AI client inspect and, if permitted, change a Grafana installation through Model Context Protocol tools. It covers dashboards, metrics, logs, alerts, incidents, and several datasource APIs, which saves an operator from pasting large Grafana responses into a chat by hand.

SetupDocsCommunityMaturity

Verdict Our mcp-grafana run installed 431 packages and passed all 10 tests, so the code is an easy trial for an existing Grafana team even though the measured build took 113 seconds. Use it when an assistant needs to move across dashboards, logs, metrics, and incident data under a service account you can restrict.

Full review →
142

system-design-primer

donnemartin/system-design-primer · ★ 370k1.2k/7d · Python

The System Design Primer is a massive, open-source knowledge repository that serves as a comprehensive guide to designing large-scale software systems. It consolidates a vast amount of information, tutorials, and diagrams from across the web into a single, structured curriculum. Its primary goal is to help software engineers prepare for the notoriously difficult system design portion of technical interviews.

SetupDocsCommunityMaturity

Verdict The System Design Primer is the definitive, free starting point for anyone serious about mastering system design for technical interviews. While its sheer size can be daunting and its maintenance seems stretched thin, no other resource offers this breadth and depth of organized knowledge at zero cost.

Full review →
143

developer-roadmap

nilbuild/developer-roadmap · ★ 367k686/7d · TypeScript

developer-roadmap is the content repository behind roadmap.sh, a collection of visual learning paths for software roles and technologies. It helps learners decide what to study next, then links each topic to short explanations and outside resources.

SetupDocsCommunityMaturity

Verdict Our run installed 31 packages in 15 seconds, but the repository exposed no build or test target, which fits a content mirror better than an application. Use roadmap.sh to choose a direction and discuss priorities, then pair it with a course, projects, and feedback.

Full review →
144

project-based-learning

practical-tutorials/project-based-learning · ★ 283k800/7d · Python

Project Based Learning is a curated directory of free tutorials that teach programming by building a working artifact. The README groups external lessons by primary language, making it a large idea shelf for learners who would rather build an interpreter, game, server, or application than follow a concept-only course.

SetupDocsCommunityMaturity

Verdict Our Project Based Learning checkout built in 3 seconds, but its 14 files exposed no test target and almost all teaching happens on external sites. Use it as a free project finder once you can judge prerequisites and tutorial age for yourself.

Full review →
145

vscode

microsoft/vscode · ★ 193k1.1k/7d · TypeScript

The `microsoft/vscode` repository is Code OSS, the MIT-licensed source base used to develop Visual Studio Code. It combines an Electron editor shell, language features, debugging, terminals, remote-development pieces, extension APIs, and bundled extensions; Microsoft's downloadable VS Code product adds separate customizations and uses a different product license.

SetupDocsCommunityMaturity

Verdict Our VS Code checkout spent 505 seconds in `npm install` before node-gyp failed on `native-keymap`, so source contribution needs more care than installing the polished desktop product. Use Code OSS when you intend to change the editor or own a distribution, and match `.nvmrc` plus every native prerequisite before debugging JavaScript.

Full review →
146

terminal

microsoft/terminal · ★ 105k55/7d · C++

Windows Terminal is Microsoft's open-source terminal application for Windows, with tabs, profiles, themes, rich text, and support for command-line programs such as PowerShell, Command Prompt, and WSL shells. The same repository also contains the Windows console host and reusable console components, so it is both an end-user app and part of the Windows command-line codebase.

SetupDocsCommunityMaturity

Verdict Our 3-CPU, 8 GB Debian sandbox could not run commit `ad06aa0` because the C++ project has no supported lab path and no Dockerfile. For Windows users, the released app is still the sensible default: installation is easy, profiles cover the shells people actually use, and development was active on August 25, 2026.

Full review →
147

storybook

storybookjs/storybook · ★ 91k40/7d · TypeScript

Storybook is a workshop for building, viewing, testing, and documenting interface components away from the rest of an application. It gives teams a shared place to inspect UI states without navigating through every product flow that normally contains them.

SetupDocsCommunityMaturity

Verdict Our test command failed in 6 seconds with MODULE_NOT_FOUND for @storybook/addon-vitest. Storybook is still the safest default here for a serious, multi-framework component program because its isolation model, addon API, documentation path, and adoption are unusually hard to match.

Full review →
148

mermaid

mermaid-js/mermaid · ★ 90k83/7d · TypeScript

Mermaid turns Markdown-like text into diagrams and charts, so architecture and process visuals can live beside the documentation they explain. It tackles diagram drift by making visuals easy to edit in source control instead of locking them inside a drawing file.

SetupDocsCommunityMaturity

Verdict Our run installed 2,209 packages, then completed both the 38-second build and 217-second test step successfully. Mermaid is the sensible default when diagrams need to live in Markdown, source control, or a JavaScript application, especially because GitHub and many other tools already understand it.

Full review →
149

hoppscotch

hoppscotch/hoppscotch · ★ 80k48/7d · TypeScript

Hoppscotch is an API client available in the browser, as desktop apps, and through a CLI. It lets developers send and organize HTTP, GraphQL, WebSocket, Socket.IO, MQTT, and event-stream requests, then share collections, environments, and tests with a team.

SetupDocsCommunityMaturity

Verdict Hoppscotch is an easy recommendation for developers who want a capable API client that works immediately in a browser and can later move to desktop or self-hosting. Its protocol range and successful lab tests are convincing, but the source install is large and the self-hosted product is a real multi-service application.

Full review →
150

awesome-react

enaqx/awesome-react · ★ 75k73/7d

Awesome React is an English-language directory of React and React Native frameworks, libraries, learning material, development tools, and example applications. It solves the first hour of discovery by putting 148 resources into one browsable README, though it does not rank or test those choices.

SetupDocsCommunityMaturity

Verdict Our 2026-09-07 sandbox found no runnable target at commit e6fd973, so awesome-react should be judged as an edited directory rather than software. Use it to assemble a candidate list when you already know how to inspect each candidate yourself.

Full review →
151

awesome-cpp

fffaraz/awesome-cpp · ★ 73k125/7d

Awesome C++ is an English-language index of C and C++ libraries, frameworks, tools, books, talks, and reference sites. It helps developers find candidates by topic, with a short description and usually a license label, then sends them to each project's own documentation.

SetupDocsCommunityMaturity

Verdict Our lab found no runnable ecosystem or Dockerfile, so Awesome C++ works as a 1,392-link discovery index and should not be read as tested dependency advice. Use it to make a candidate list, especially when you already know how to inspect a C++ project.

Full review →
152

swift

swiftlang/swift · ★ 70k11/7d · Swift

Swift is a memory-safe systems programming language with modern syntax and direct access to C and Objective-C code. This repository contains the compiler and standard library, while ordinary application developers can install a released toolchain instead of building this source tree.

SetupDocsCommunityMaturity

Verdict Our run of Swift's `cmpcodesize` utility installed 36 packages and passed 3 of 3 tests, but that small success should not be mistaken for building the 3.5-million-line compiler. Choose Swift when Apple-platform integration or Swift's language model is the deciding factor.

Full review →
153

Java

TheAlgorithms/Java · ★ 66k16/7d · Java

TheAlgorithms/Java is a large, searchable collection of algorithms and data structures written in Java for study and contribution. It helps developers inspect working implementations by topic, but the project explicitly says its educational code may be less efficient than Java's standard library.

SetupDocsCommunityMaturity

Verdict Our Java run passed all 29,285 tests in 121 seconds, making TheAlgorithms/Java a strong code-reading resource for learners who want examples they can execute. Use it to study, compare approaches, and practice contributions.

Full review →
154

alacritty

alacritty/alacritty · ★ 66k61/7d · Rust

Alacritty is a GPU-rendered terminal emulator for BSD, Linux, macOS, and Windows. It concentrates on drawing a terminal quickly and delegates tabs, panes, and session management to a window manager or terminal multiplexer.

SetupDocsCommunityMaturity

Verdict Our Alacritty build passed in 66 seconds and 219 of 220 tests passed, with one Unix password-entry test failing on `pw not found`. Use it if you already want tmux or your window manager to own panes and tabs, and you value a small terminal surface over built-in conveniences.

Full review →
155

traefik

traefik/traefik · ★ 65k58/7d · Go

Traefik is a reverse proxy and load balancer that watches Docker, Kubernetes, and other infrastructure sources, then updates routes as services change. It removes much of the manual proxy configuration from environments where containers appear, move, or disappear frequently.

SetupDocsCommunityMaturity

Verdict Our Traefik build passed in 311 seconds, but 3 of 50 Go test groups failed after a 597-second run without a specific error in the supplied tail. Use Traefik when Docker or Kubernetes route churn makes dynamic discovery worth the control-plane work.

Full review →
156

FFmpeg

FFmpeg/FFmpeg · ★ 64k199/7d · C

FFmpeg is the source project behind command-line tools and C libraries for decoding, encoding, filtering, inspecting, playing, and streaming audio, video, subtitles, and metadata. It solves media plumbing problems at a depth that most friendlier converters reach by calling FFmpeg underneath.

SetupDocsCommunityMaturity

Verdict Our FFmpeg source install failed after 10 seconds because `nasm` was missing or too old, so the 1,850,143-line checkout never reached build or tests. Most users should install a trusted packaged binary and inspect its configuration before writing commands against it.

Full review →
157

rustlings

rust-lang/rustlings · ★ 64k45/7d · Rust

Rustlings is a set of small Rust exercises plus a terminal program that checks your work as you edit. It gives beginners a short feedback loop for compiler errors, tests, hints, and progress while they read the official Rust Book.

SetupDocsCommunityMaturity

Verdict Our Rustlings run installed 67 packages and built in 54 seconds combined, but only 20 of 26 maintainer tests passed, so beginners can try it easily while CI users should reproduce the failures first. It remains an excellent companion to the Rust Book because the edit, compile, hint, and retry loop keeps abstract ownership rules tied to code.

Full review →
158

gitea

go-gitea/gitea · ★ 58k77/7d · Go

Gitea is a self-hosted software development service for Git repositories, code review, issues, projects, packages, wikis, and CI jobs compatible with GitHub Actions. It gives a team one place to run its development work without handing repository control to a hosted forge.

SetupDocsCommunityMaturity

Verdict Our Gitea checkout pulled 913 packages and occupied 579 MB, yet exposed no build or test target to the harness, so source contributors should budget for the project's separate Go and frontend workflow. Use Gitea when you want a capable private forge without GitLab's wider operating footprint and you have an administrator for storage, upgrades, mail, and runners.

Full review →
159

lazydocker

jesseduffield/lazydocker · ★ 53k70/7d · Go

Lazydocker is a terminal interface for inspecting and controlling Docker containers, images, volumes, and Compose services. It replaces a loop of `docker ps`, log commands, restarts, and cleanup commands with panels and keyboard shortcuts in one terminal.

SetupDocsCommunityMaturity

Verdict Our lazydocker run built in 24 seconds and passed all 6 measured tests in 11 seconds, which makes it the cleanest trial in this group. Use it for fast, single-user Docker troubleshooting from a terminal, especially when logs and restarts keep sending you across shells.

Full review →
160

dbeaver

dbeaver/dbeaver · ★ 52k64/7d · Java

DBeaver is a desktop database client for browsing schemas, writing SQL, editing data, moving records, and administering many kinds of databases from one application. The free Community edition covers JDBC databases, while paid editions add several non-JDBC sources and extra tooling.

SetupDocsCommunityMaturity

Verdict Our DBeaver source build stopped after 4 seconds because Maven could not resolve its snapshot parent POM, although the install step finished in 6 seconds. Use the prebuilt Community release if you want one capable desktop client for many JDBC databases; it is a much safer entry point than treating this as an ordinary Maven project.

Full review →
161

DefinitelyTyped

DefinitelyTyped/DefinitelyTyped · ★ 51k10/7d · TypeScript

DefinitelyTyped is the community repository behind npm's `@types` packages, which add TypeScript declarations to JavaScript libraries that do not ship their own. Most developers use one published package such as `@types/node`; the huge monorepo is mainly for people reviewing or correcting those declarations.

SetupDocsCommunityMaturity

Verdict Our DefinitelyTyped checkout reached 62,951 files and npm stopped after 70 seconds because the monorepo requires pnpm. Consumers should still use a focused `@types` package when a JavaScript dependency lacks declarations.

Full review →
162

jadx

skylot/jadx · ★ 50k76/7d · Java

JADX is a desktop and command-line decompiler that turns Android APK and Dex bytecode into Java-like source you can search and inspect. It also decodes the app manifest and resources, helping developers and security analysts understand an Android package when the original source is unavailable.

SetupDocsCommunityMaturity

Verdict Our JADX run completed its 149-second install, 96-second build, and 224-second test step without failure, making v1.5.6 an easy recommendation for first-pass Android code inspection. Use its output as an informed reconstruction, then check suspicious control flow against smali or another tool.

Full review →
163

supervision

roboflow/supervision · ★ 50k360/7d · Python

Supervision is a Python toolkit for the work around a computer vision model: storing detections, drawing annotations, tracking objects, measuring geometry, counting zones, and converting datasets. It lets an application use outputs from different detection or segmentation models without rewriting the same post-processing code for each one.

SetupDocsCommunityMaturity

Verdict Our Supervision run installed 50 packages and built in 7 seconds, while the test step ended with 2 `sklearn` collection errors after logging 3,517 passes. That is good enough to justify a trial for model-neutral vision plumbing, especially annotation, tracking, geometry, and dataset conversion.

Full review →
164

GitNexus

abhigyanpatwari/GitNexus · ★ 47k210/7d · TypeScript

GitNexus turns a codebase into a local knowledge graph that coding agents can query for symbols, call paths, dependencies, and change impact. It connects that graph to Claude Code, Codex, Cursor, and other clients through MCP, with a browser explorer for visual inspection.

SetupDocsCommunityMaturity

Verdict Our GitNexus install took 15 seconds and left 205 packages using 93 MB, but the repository root exposed no build or test target for us to verify. Use it as a local second opinion when an existing coding agent needs graph-backed context across a large codebase.

Full review →
165

serverless

serverless/serverless · ★ 47k1/7d · JavaScript

Serverless Framework is a command-line tool that turns a YAML service description into AWS Lambda code and the cloud resources around it. Version 4 covers deployment, logs, local-to-cloud development, multiple function languages, and an extension system, with AWS now the clear focus.

SetupDocsCommunityMaturity

Verdict Our Node 22 run installed Serverless in 431 seconds, and 3,757 of 3,758 tests passed before `zip-service.test.js` left the command at exit 1. Use V4 when AWS Lambda and CloudFormation are already the team's standard and the Serverless login and license split have been approved.

Full review →
166

yazi

sxyazi/yazi · ★ 42k129/7d · Rust

Yazi is a keyboard-driven file manager that runs inside a terminal on Linux, macOS, Windows, and Android. It gives terminal users fast directory browsing, rich file previews, bulk operations, tabs, and Lua customization without sending them back to a desktop file manager.

SetupDocsCommunityMaturity

Verdict We built Yazi successfully in 264 seconds, but `cargo test` ran 0 tests, so the project is easier to build than its green test result is to trust. It is still the terminal file manager I would try first for rich previews, background file work, and deep customization.

Full review →
167

stats

exelban/stats · ★ 42k · Swift

Stats is a free macOS menu bar app for watching CPU, GPU, memory, disks, network traffic, battery data, sensors, Bluetooth devices, and clocks. It solves the everyday problem of checking system activity without leaving a monitor window open.

SetupDocsCommunityMaturity

Verdict We did not run Stats in our 3-CPU Debian sandbox because commit 85543cb is a Swift macOS app, an unsupported ecosystem there, and the repository has no Dockerfile. It is still an easy recommendation for Mac users who want a free, configurable menu monitor and can verify it on their own stable macOS release.

Full review →
168

omarchy

basecamp/omarchy · ★ 41k1.5k/7d · Shell

Omarchy is an opinionated Arch Linux distribution that turns a bare PC into a keyboard-driven Hyprland workstation with a themed desktop and a large set of developer tools. It solves the tedious job of assembling, styling, and maintaining an Arch-based coding environment, but it does so by making many choices for you.

SetupDocsCommunityMaturity

Verdict Omarchy is compelling if its screenshots, shortcuts, and chosen applications already resemble the workstation you want. It saves weeks of desktop assembly and now offers a much more coherent shell, but Quattro is a fresh architectural rewrite with serious open upgrade and locking reports.

Full review →
169

manim

ManimCommunity/manim · ★ 41k97/7d · Python

Manim Community is a Python engine for making explanatory animations, especially equations, graphs, geometry, and other mathematical ideas. You describe each scene in code, then Manim renders it as video or an image, which makes precise transformations easier to repeat than manual timeline editing.

SetupDocsCommunityMaturity

Verdict Our Manim install and 9-second build passed, but its tests stopped after 8 seconds because Python could not import `cairo`. Use v0.21.0 for code-driven mathematical animation if you can own the native rendering stack and pin scene projects through the current refactor.

Full review →
170

llvm-project

llvm/llvm-project · ★ 40k142/7d · LLVM

LLVM Project is the source tree for a family of compiler and toolchain components, including LLVM's optimizer and code generator, Clang, libc++, and the LLD linker. It is for building compilers, language runtimes, analysis tools, and complete native-code toolchains rather than adding one small library to an application.

SetupDocsCommunityMaturity

Verdict Our 2,270.2 MB LLVM checkout installed 35 Python packages, but the 9-second harness build failed before producing a native toolchain, which shows how poorly a generic project runner fits this repository. Use LLVM when you need its compiler libraries, targets, or one of its major subprojects and can budget for a selective CMake build.

Full review →
171

awesome-copilot

github/awesome-copilot · ★ 39k264/7d · JavaScript

Awesome Copilot is a community catalog of reusable agents, instructions, skills, hooks, workflows, plugins, and API recipes for GitHub Copilot. It gives developers searchable starting points for specific jobs instead of requiring every customization to be written from scratch.

SetupDocsCommunityMaturity

Verdict Our Awesome Copilot run installed 150 packages and built in 13 seconds with 0 npm audit findings, but it exposed no test target for the catalog's promised behavior. It is the best first search stop for GitHub Copilot users who know the job they want to improve.

Full review →
172

lerna

lerna/lerna · ★ 36k1/7d · TypeScript

Lerna is a command-line tool for running tasks, versioning packages, and publishing many JavaScript or TypeScript packages kept in one repository. It solves the coordination work around package order, changed-project selection, release versions, and npm publishing while leaving dependency installation to npm, Yarn, pnpm, or Bun.

SetupDocsCommunityMaturity

Verdict Our Lerna run installed 1,847 packages and occupied 685 MB, then built successfully and passed all 28 tests, so it is dependable to verify but expensive to carry. Use it when a JavaScript monorepo needs package-aware task runs and npm releases in one familiar CLI.

Full review →
173

OCRmyPDF

ocrmypdf/OCRmyPDF · ★ 35k · Python

OCRmyPDF is a command-line tool that adds a searchable, copyable text layer to scanned PDFs while preserving the page images. It solves the practical gap between a PDF that merely looks readable and one that can be searched, indexed, archived, or processed by other software.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 6 seconds, but the repository's test command stopped after 8 seconds because pytest did not recognize `-n`. OCRmyPDF is still the strongest default for technical users who need searchable, archival PDFs without sending documents to a hosted service.

Full review →
174

raylib

raysan5/raylib · ★ 35k98/7d · C

raylib is a C library for making games, visual tools, and small graphics applications by writing code directly. It bundles windowing, input, drawing, audio, math, and asset loading behind a compact API, with examples doing much of the teaching.

SetupDocsCommunityMaturity

Verdict Our raylib configuration failed in 12 seconds because the fresh Debian container lacked X11 development files, so source setup on Linux is simple only after the platform packages are in place. Once that requirement is understood, raylib is an excellent first choice for learning graphics through code and for small games that do not need an editor or complex text layout.

Full review →
175

k9s

derailed/k9s · ★ 35k54/7d · Go

K9s is a full-screen terminal interface for viewing and operating Kubernetes clusters through the credentials in your kubeconfig. It turns repeated kubectl work into keyboard-driven resource lists, logs, YAML views, filters, port forwarding, shells, edits, deletes, and custom commands.

SetupDocsCommunityMaturity

Verdict Our K9s build completed in 241 seconds, but 3 of 21 Go packages failed their tests, so commit 24cf462 is usable for evaluation without earning a clean engineering bill. K9s remains the strongest terminal choice for experienced Kubernetes operators who value speed, filters, logs, and extensibility.

Full review →
176

backstage

backstage/backstage · ★ 34k44/7d · TypeScript

Backstage is an open-source framework for building a company developer portal around a shared software catalog. It gives engineers one place to find services, owners, documentation, and approved project templates instead of hunting across separate infrastructure tools.

SetupDocsCommunityMaturity

Verdict Our test run hit the 900-second cap after a 183-second install. Backstage is a strong choice when a staffed platform team needs a customizable catalog, templates, and documentation hub, but it is not a ready-made portal you install and forget.

Full review →
177

learnGitBranching

pcottle/learnGitBranching · ★ 34k17/7d · JavaScript

LearnGitBranching is a browser-based Git simulator that turns commands into an animated commit graph and a series of short challenges. It helps developers understand branches, references, rebases, cherry-picks, and remotes without risking a real repository.

SetupDocsCommunityMaturity

Verdict Our sandbox installed 598 packages in 25 seconds, built the app in 12 seconds, and passed its tests in 8 seconds, so LearnGitBranching is straightforward to verify even with a sizable dependency tree. Use it when someone knows Git commands but cannot yet picture what they do to commits and references.

Full review →
178

BetterDisplay

waydabber/BetterDisplay · ★ 34k

BetterDisplay is a macOS menu-bar app for fixing scaling, brightness, color, virtual-screen, and multi-monitor annoyances. Its public GitHub repository contains only an English README, with no detected programming language or license, so readers can inspect the product documentation and issue tracker but not the app's source.

SetupDocsCommunityMaturity

Verdict Our sandbox could not run BetterDisplay because the repository exposes no supported language and has no Dockerfile. For Mac owners fighting HiDPI, DDC, virtual-screen, or arrangement limits, the documented feature set earns a 14-day Pro trial on the exact hardware that will use it.

Full review →
179

WxJava

binarywang/WxJava · ★ 33k · Java

WxJava is a Chinese-first Java SDK for adding WeChat services to a backend. Its main README and linked wiki are Chinese, and the repository has no parallel English README or English guide. It wraps official accounts, mini programs, payments, corporate WeChat, the open platform, and Channels or WeChat Store, but it is not a ready-made web application or mobile client.

SetupDocsCommunityMaturity

Verdict Our WxJava sandbox installed in 63 seconds, built in 81 seconds, and passed all 272 Maven tests, which makes this large SDK practical to trial. Use it for a Java backend that must span several WeChat services and has Chinese-reading maintainers.

Full review →
180

podman

podman-container-tools/podman · ★ 33k61/7d · Go

Podman is a command-line engine for building, running, and managing OCI containers, images, volumes, networks, and pods. It offers familiar Docker-style commands without a permanent manager daemon, supports rootless containers on Linux, and uses a managed virtual machine on macOS and Windows.

SetupDocsCommunityMaturity

Verdict Our Podman checkout installed 35 Python packages and built in 22 seconds combined, but all 47 REST tests errored during setup because `bin/podman` was absent. That result makes the source checkout a poor generic Python install, not a rejection of Podman as a container engine.

Full review →
181

refined-github

refined-github/refined-github · ★ 32k74/7d · TypeScript

Refined GitHub is a browser extension that changes GitHub's web interface to remove annoyances and add shortcuts, context, and safer defaults. It improves repository pages, pull requests, comments, notifications, releases, and other views without asking a team to replace GitHub.

SetupDocsCommunityMaturity

Verdict Our Refined GitHub run built in 19 seconds and all 564 executed Vitest checks passed, but the full test command exited 1 on one ESLint rule. Install the store version if GitHub is where you spend hours each day and small interface fixes save repeated clicks.

Full review →
182

skills

vercel-labs/skills · ★ 32k891/7d · TypeScript

Skills is an English-language command-line manager for finding, installing, updating, using, and removing instruction packages across Claude Code, Codex, Cursor, OpenCode, and dozens of other coding agents. It solves the incompatible-folder problem by discovering `SKILL.md` files once and placing them where each supported agent expects them.

SetupDocsCommunityMaturity

Verdict Our Skills run installed 126 packages in 5 seconds, then passed its 9-second build and 32-second test step, making it a credible cross-agent installer with one serious current caution. Use it for add, list, use, and carefully scoped updates across mixed agent setups.

Full review →
183

k6

grafana/k6 · ★ 31k · Go

k6 is a command-line load-testing tool that lets developers describe traffic and performance expectations in JavaScript while its Go-based engine generates the load. It helps teams find slow responses, capacity limits, and performance regressions before users encounter them.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 109 seconds, but tests ended at 133 passed and 4 failed. k6 is still an excellent default for engineering teams that want expressive, version-controlled performance tests across several protocols, especially when Grafana is already part of the stack.

Full review →
184

nuclei

projectdiscovery/nuclei · ★ 31k146/7d · Go

Nuclei is a command-line vulnerability scanner that sends checks described in YAML templates against websites, APIs, networks, DNS, and other targets. It lets security teams reuse community detections and write their own repeatable checks instead of building a scanner for each vulnerability.

SetupDocsCommunityMaturity

Verdict Our Nuclei run installed 758 packages and built in 269 seconds, but 3 of 143 tests failed with no cause visible in the supplied tail. Authorized security teams should still trial it because the template model makes focused, repeatable checks easy to inspect.

Full review →
185

openssl

openssl/openssl · ★ 31k17/7d · C

OpenSSL is a C toolkit for securing network connections with TLS, DTLS, and QUIC, plus a general-purpose cryptography library. It gives operating systems, servers, and application developers the protocol code, certificate tooling, and cryptographic primitives needed to protect data in transit.

SetupDocsCommunityMaturity

Verdict Our build completed in 69 seconds, and all 1,046 tests passed. The measurement targeted the recorded Rust subproject, so it is encouraging integration evidence rather than a clean native OpenSSL build verdict.

Full review →
186

opentofu

opentofu/opentofu · ★ 30k55/7d · Go

OpenTofu is an open-source command-line tool for describing cloud and data-center infrastructure in code, previewing changes, and applying them. It gives Terraform users a community-governed implementation with its own registry while preserving a familiar configuration workflow.

SetupDocsCommunityMaturity

Verdict Our OpenTofu build succeeded in 181 seconds, but 3 of 130 Go package tests failed, so the CLI is credible for an evaluated migration rather than a blind swap. Use it if community governance and an open registry matter, and your team can compare plans, protect state, and test providers before applying.

Full review →
187

micro

micro-editor/micro · ★ 30k45/7d · Go

micro is a terminal text editor built for people who find nano approachable but want tabs, splits, mouse support, multiple cursors, plugins, and better defaults. It ships as a single Go binary and is meant to work both as an everyday terminal editor and as a practical tool over SSH.

SetupDocsCommunityMaturity

Verdict Our build finished in 24 seconds and all 12 measured tests passed, which makes micro an easy editor to trial and a credible tool to keep on remote machines. Pick it when nano feels limiting but Vim's interaction model feels like a tax.

Full review →
188

nx

nrwl/nx · ★ 29k25/7d · TypeScript

Nx is a build system and workspace manager for monorepos, especially TypeScript codebases that have grown into many applications and packages. It discovers project relationships, caches task output, runs affected work, and adds generators and CI coordination around the tools a team already uses.

SetupDocsCommunityMaturity

Verdict Our Nx checkout installed 4,313 packages, used 3,120 MB, then failed both build and tests because `dotnet` was absent, so contributing to Nx requires a genuinely polyglot toolchain. Using Nx inside an application repo is much lighter and worthwhile once task graphs, affected runs, or CI caching save measurable time.

Full review →
189

wezterm

wezterm/wezterm · ★ 29k89/7d · Rust

WezTerm is a GPU-rendered terminal emulator and terminal multiplexer for Linux, macOS, Windows, FreeBSD, and NetBSD. It combines ordinary terminal windows with panes, tabs, searchable scrollback, Lua configuration, SSH connections, serial-port access, and local or remote persistent sessions.

SetupDocsCommunityMaturity

Verdict Our WezTerm checkout pulled 790 packages, then both build and test stopped because the Debian sandbox lacked `x11-xcb`, so source setup is not a Cargo-only job. Install a published package if you want a programmable terminal with panes, SSH, graphics, and cross-platform configuration.

Full review →
190

pydantic

pydantic/pydantic · ★ 29k · Python

Pydantic turns Python type hints into runtime checks, parsed model objects, serialization, and JSON Schema. It is useful at API, configuration, and data-loading boundaries where outside input must become predictable Python values.

SetupDocsCommunityMaturity

Verdict Our Pydantic run installed 39 packages in 89 seconds, but the build and tests both failed in 2 seconds, so this source checkout is not a clean recommendation for contributors. Application teams should still shortlist Pydantic when typed models, parsing, and JSON Schema need to share one definition.

Full review →
191

trufflehog

trufflesecurity/trufflehog · ★ 28k205/7d · Go

TruffleHog finds credentials in Git history, files, container images, cloud storage, collaboration systems, and build services. It can classify many findings by provider, check whether credentials are still live, and gather access details for selected credential types.

SetupDocsCommunityMaturity

Verdict Our TruffleHog build succeeded, but 7 of 220 tests failed after a 521-second run, so pin and qualify the exact version before making it a mandatory gate. It remains a strong choice when live credential verification and broad source coverage materially improve incident response.

Full review →
192

LazyVim

LazyVim/LazyVim · ★ 27k45/7d · Lua

LazyVim is an English-documented Neovim setup that turns the editor into a configured coding environment while leaving its Lua settings open to change. It supplies plugin choices, editor defaults, language extras, and update machinery so users do not have to assemble everything from scratch.

SetupDocsCommunityMaturity

Verdict Our sandbox could not execute commit 9997009 because the harness supports no Lua ecosystem and the repository has no Dockerfile, so we have no install or test result to offset the documentation. LazyVim v16.0.1 is still one of the easier ways for an experienced terminal user to adopt a maintained Neovim setup without giving up Lua overrides.

Full review →
193

Wox

Wox-launcher/Wox · ★ 27k58/7d · Go

Wox is an open-source desktop launcher for macOS, Linux, and Windows that puts app launching, file finding, and common actions behind one keyboard-driven search box. It is meant to cut down the time spent moving through menus and separate utilities, while plugins add tasks beyond the built-in search experience.

SetupDocsCommunityMaturity

Verdict Our install and build succeeded in 69 seconds total, but the test step executed 0 Go tests. Wox is an appealing choice if you want one native, keyboard-first launcher across 3 desktop platforms and value an open plugin system.

Full review →
194

glow

charmbracelet/glow · ★ 27k88/7d · Go

Glow is a terminal Markdown reader that renders local files, standard input, web URLs, and repository pages without opening a browser. Run it with no arguments and it becomes a text interface for finding and reading Markdown below the current directory.

SetupDocsCommunityMaturity

Verdict Our Glow run installed 98 packages, built successfully, and passed 2 of 2 tests in 122 seconds total, making it an easy trial for anyone who lives in a terminal. Use it for quick README browsing and styled pipeline output.

Full review →
195

hashcat

hashcat/hashcat · ★ 27k38/7d · C

Hashcat is a command-line password recovery and auditing tool that tests candidate passwords against hashes you already possess. It runs on GPUs and CPUs, covers hundreds of hash formats, and lets authorized security teams check whether captured or exported password data falls to realistic wordlists and rules.

SetupDocsCommunityMaturity

Verdict Our Hashcat run built `Rust/hashcat-sys` in 15 seconds and passed its 2 tests in 6 seconds, but it did not exercise the 423,663-line native engine or any GPU backend. Use the official v7.1.2 package for authorized offline recovery when you can verify the hash mode and compute runtime on the exact host.

Full review →
196

rollup

rollup/rollup · ★ 26k · JavaScript

Rollup combines JavaScript modules into files that browsers, Node.js, and other tools can consume. It solves the practical problem of writing a codebase with standard ES modules while shipping optimized bundles in modern or legacy formats, including CommonJS, UMD, AMD, and IIFE.

SetupDocsCommunityMaturity

Verdict Our installation stopped after 608 seconds at a wasm-pack v0.15.0 download that timed out. Rollup remains a strong choice for JavaScript library output because its purpose is narrow, its format support is explicit, and its ES module analysis directly serves package publishing.

Full review →
197

nixpkgs

NixOS/nixpkgs · ★ 26k79/7d · Nix

Nixpkgs is the package collection behind the Nix package manager and the module base used to build NixOS. It describes how software and whole systems are assembled reproducibly, while Hydra builds accepted expressions and publishes successful artifacts to the shared binary cache.

SetupDocsCommunityMaturity

Verdict Our check of `ci/github-script/` took 11 seconds, installed 0 packages, and found 5 known vulnerabilities, but it did not build Nixpkgs or run NixOS tests. Use Nixpkgs when reproducible packages and declarative systems justify learning Nix and pinning a moving collection.

Full review →
198

ILSpy

icsharpcode/ILSpy · ★ 26k37/7d · C#

ILSpy turns compiled .NET assemblies back into readable C# so developers can understand libraries when source code is unavailable. It combines a cross-platform desktop browser with command-line, PowerShell, Visual Studio, Visual Studio Code, and reusable-library options.

SetupDocsCommunityMaturity

Verdict Our scoped npm install completed in 26 seconds with 396 packages and 0 known vulnerabilities. ILSpy is the default open-source recommendation for serious .NET decompilation because it pairs a mature engine with desktop, CLI, editor, PowerShell, and library frontends.

Full review →
199

cilium

cilium/cilium · ★ 25k44/7d · Go

Cilium is a networking, security, and traffic-observability layer built mainly for Kubernetes on Linux. It uses eBPF programs in the kernel to connect workloads, enforce identity-aware policies, balance services, and show operators why traffic was allowed or dropped.

SetupDocsCommunityMaturity

Verdict Our Cilium build took 438 seconds and 209 of 210 tests passed, so skilled Kubernetes teams have a credible codebase here, with one unresolved failure in our sandbox. Put Cilium on the shortlist when policy, service networking, and Hubble visibility solve problems your team already has.

Full review →
200

slint

slint-ui/slint · ★ 24k57/7d · Rust

Slint is a declarative GUI toolkit for native applications on embedded hardware, desktop systems, mobile devices, and the web. Developers describe the interface in Slint's own markup language, then connect it to business logic written in Rust, C++, JavaScript, or Python.

SetupDocsCommunityMaturity

Verdict Slint is one of the strongest open-source choices for a team that needs the same native UI model across Rust or C++, desktop, and constrained devices. The stable 1.x core, renderer choices, tooling, and active releases justify a serious prototype.

Full review →
201

ruby

ruby/ruby · ★ 24k34/7d · Ruby

ruby/ruby is the source tree for the Ruby programming language and its main CRuby interpreter. Ruby is a dynamic, object-oriented language used for web applications, scripts, text processing, and system tasks, while this repository is mainly for interpreter contributors and people building Ruby from source.

SetupDocsCommunityMaturity

Verdict Our Cargo-oriented run built in 4 seconds but discovered 0 tests, so it says nothing about whether a full CRuby checkout passes its documented Make suites. Use ruby/ruby when you need the reference implementation, a specific 4.0.6 fix, or a place to contribute to the language.

Full review →
202

magic-wormhole

magic-wormhole/magic-wormhole · ★ 23k12/7d · Python

Magic Wormhole is a command-line tool and Python library for sending files, directories, or short text between two computers. The sender reads out a short one-use code, the receiver enters it, and the two clients arrange an encrypted direct or relayed transfer without exchanging accounts or SSH keys.

SetupDocsCommunityMaturity

Verdict Our Magic Wormhole run installed 76 packages in 51 seconds and passed 462 tests with 0 failures, making it an easy recommendation for occasional person-to-person transfers. Keep the source copy until the receiver verifies a large directory, because issue #722 documents a completed-looking 31.53 GB attempt that left no files after interruption.

Full review →
203

unleashed-firmware

DarkFlippers/unleashed-firmware · ★ 22k61/7d · C

Unleashed Firmware is an unofficial Flipper Zero firmware fork with English documentation, extra Sub-GHz controls and protocols, NFC changes, bundled community apps, and interface tweaks. It is for owners who want more radio and device functionality than the official firmware exposes and who accept responsibility for custom firmware.

SetupDocsCommunityMaturity

Verdict Our Unleashed checkout occupied 199.7 MB, and pytest found 0 tests in `lib/nanopb`, so the 3-second nested build is not evidence that the firmware passed its test suite. Flipper Zero owners who want Unleashed's extra radio protocols and apps should try the stable release, keep a recovery path, and test the exact functions they depend on.

Full review →
204

nnn

jarun/nnn · ★ 22k7/7d · C

nnn is a terminal file manager for browsing, selecting, renaming, moving, and previewing files without leaving the shell. It keeps the core program small, then hands jobs such as remote mounts, rich previews, and uploads to optional plugins and command-line programs.

SetupDocsCommunityMaturity

Verdict Our nnn run installed in 6 seconds and built in 21 seconds, but its test step stopped with `no test target`, so the program is easy to compile while the checkout gives a new contributor no passing test command. Use it if you want a fast keyboard file browser and are happy to compose previews and integrations from local tools.

Full review →
205

desktop

desktop/desktop · ★ 22k18/7d · TypeScript

GitHub Desktop is a graphical Git client for macOS and Windows, built by GitHub with Electron, TypeScript, and React. It makes common branch, commit, push, pull, and pull-request work visible for people who would rather not perform every action in a terminal.

SetupDocsCommunityMaturity

Verdict Our GitHub Desktop install pulled 1,123 packages and occupied 1,232 MB, then its 124-second test run failed with only repeated `test failed` lines in the supplied tail. The packaged client remains a good choice for macOS or Windows users who want GitHub-focused visual commits and branch work.

Full review →
206

lo

samber/lo · ★ 21k6/7d · Go

Lo is a Go utility library that applies familiar operations such as map, filter, group, and reduce to typed slices and maps. It cuts repetitive loops from application code and also includes separate packages for lazy iteration, in-place changes, and parallel callbacks.

SetupDocsCommunityMaturity

Verdict Our Lo run installed 16 packages in 9 seconds, built in 28 seconds, and passed all 12 `go test` checks in 26 seconds. Use its stable v1 core when repeated collection loops are obscuring ordinary application code, but keep direct loops in measured hot paths.

Full review →
207

roslyn

dotnet/roslyn · ★ 21k15/7d · C#

Roslyn is Microsoft's open-source implementation of the C# and Visual Basic compilers, plus APIs that let programs inspect, understand, rewrite, and compile source code. It is the foundation for analyzers, code fixes, refactoring tools, editor intelligence, and language services across the .NET ecosystem.

SetupDocsCommunityMaturity

Verdict Our sandbox did not run commit db9d3ac because C# is unsupported and the repository supplied no Dockerfile, so source-build ease remains unverified here. Use Roslyn's packages when your tool needs correct C# or Visual Basic semantics, diagnostics, or code changes; writing a substitute parser is rarely sensible.

Full review →
208

crystal

crystal-lang/crystal · ★ 20k · Crystal

Crystal is a compiled programming language that pairs Ruby-like syntax with inferred static types and native code generated through LLVM. It includes compile-time macros, direct C bindings, a standard library, and the Shards dependency tool. The README explicitly says Ruby compatibility is not a goal, so this is its own language rather than a Ruby implementation.

SetupDocsCommunityMaturity

Verdict Our lab did not run commit 59835a3 because the harness had no supported Crystal ecosystem and the repository supplied no Dockerfile. Crystal is worth a greenfield trial for developers who want Ruby-shaped code, static inference, native binaries, and C interop, provided they can pin the compiler toolchain and check minor-release notes.

Full review →
209

TranslucentTB

TranslucentTB/TranslucentTB · ★ 20k61/7d · C++

TranslucentTB is a Windows utility that changes the taskbar's color, transparency, blur, and acrylic effect. It can switch appearances when a window is maximized or when Start, Search, or Task View opens, solving a small personalization problem without replacing the whole Windows shell.

SetupDocsCommunityMaturity

Verdict Our lab did not run TranslucentTB because the C++ repository has no supported harness ecosystem and no Dockerfile, so there is no measured install, build, or test outcome to lean on. For Windows 10 or 11 users who want only taskbar appearance controls, the Store release is easy enough to try and remove.

Full review →
210

validator

go-playground/validator · ★ 20k19/7d · Go

Go Playground Validator checks Go structs and individual values against rules written as tags or registered functions. It handles common formats, comparisons between fields, nested collections, custom types, and translated error messages, saving teams from rewriting the same request checks.

SetupDocsCommunityMaturity

Verdict Our Validator run installed 18 packages and passed all 50 tests after a 24-second build, making v10.30.4 an easy default for tag-friendly Go services. Use it when broad built-in rules, nested collection checks, and Gin compatibility outweigh the risks of runtime tag parsing.

Full review →
211

phpunit

sebastianbergmann/phpunit · ★ 20k4/7d · PHP

PHPUnit is an xUnit-style testing framework and command-line test runner for PHP. It gives developers assertions, fixtures, test doubles, data providers, and machine-readable reports for checking PHP code before it ships.

SetupDocsCommunityMaturity

Verdict Our sandbox produced no install, build, or test result for commit 2c534d3 because it did not support PHP and the repository had no Dockerfile. That evidence gap prevents an execution-based setup endorsement, although PHPUnit 13.3.2 remains the sensible default for a PHP 8.4 project that wants conventional tests and mature documentation.

Full review →
212

wasmtime

bytecodealliance/wasmtime · ★ 19k23/7d · Rust

Wasmtime is a standalone and embeddable runtime for running WebAssembly outside a browser. It uses Cranelift to compile modules and components, exposes WASI for host interaction, and has maintained bindings for Rust, C, C++, Python, .NET, Go, and Ruby.

SetupDocsCommunityMaturity

Verdict Our Wasmtime build succeeded after 389 seconds, but the 286-second test run failed when `wasm32-unknown-unknown` was unavailable to a test-program build. Wasmtime is the first runtime we would evaluate for a serious server-side WebAssembly host because its security process, WASI work, bindings, and release activity are unusually visible.

Full review →
213

katana

projectdiscovery/katana · ★ 18k62/7d · Go

Katana is a command-line crawler and Go library for finding URLs, forms, scripts, and network endpoints across a website. It gives security teams and automation pipelines scoped, rate-limited crawling for plain HTTP pages and JavaScript-heavy applications.

SetupDocsCommunityMaturity

Verdict Our Katana run installed 357 packages, built in 111 seconds, and passed all 46 Go tests, so it is a credible default crawler for security automation. Use it when scope rules, structured output, and optional browser execution save you from maintaining a custom spider.

Full review →
214

grype

anchore/grype · ★ 13k23/7d · Go

Grype is a command-line vulnerability scanner for container images, local filesystems, and software bills of materials. It matches operating-system and language packages against vulnerability data, then adds EPSS, KEV, risk scoring, and OpenVEX context to help teams sort findings.

SetupDocsCommunityMaturity

Verdict Our Grype run built in 134 seconds, but 4 of 98 test packages failed in integration, so adoption should include a pinned test image and checks for scanner drift. Use it for a fast, scriptable first pass across images, filesystems, and SBOMs, especially beside Syft.

Full review →
215

quiche

cloudflare/quiche · ★ 12k6/7d · Rust

quiche is Cloudflare's Rust implementation of the QUIC transport protocol and HTTP/3, with a thin C interface for other languages. It manages packets and connection state, while your application supplies the sockets, event loop, timers, and production server design.

SetupDocsCommunityMaturity

Verdict Our quiche run built in 216 seconds and passed all 1,386 tests, so experienced network teams have a well-tested base for custom QUIC or HTTP/3 work. Use it when owning packet I/O, timers, pacing, and connection policy is part of the job.

Full review →
216

resty

go-resty/resty · ★ 12k10/7d · Go

Resty is a Go HTTP client that turns common API work into chained method calls instead of repeated request setup and response plumbing. It handles bodies, authentication, retries, tracing, middleware, server-sent events, and other client concerns while still using Go's HTTP stack underneath.

SetupDocsCommunityMaturity

Verdict Our Resty run installed 6 packages, built in 27 seconds, and passed 2 of 2 Go test results, making it the easiest repository in this group to verify. Use v2 for a final release today, or trial v3 when its circuit breaker, hedging, load balancing, or SSE support saves real application code.

Full review →
217

bore

ekzhang/bore · ★ 11k5/7d · Rust

Bore is a small Rust command-line tool that exposes a local TCP port through a remote server. It solves the basic NAT and firewall problem for developers who need someone else to reach a service running on their laptop or private network.

SetupDocsCommunityMaturity

Verdict Our Bore run installed 95 packages and passed all 24 tests, making it an easy choice for a temporary TCP tunnel that you are willing to supervise. Use it for short-lived developer access or behind your own server controls.

Full review →
218

buildkit

moby/buildkit · ★ 10k19/7d · Go

BuildKit is the build engine used by modern Docker builds, and it can also run as a standalone daemon with its own client. It turns Dockerfiles or other frontends into a dependency graph, executes independent work concurrently, and moves build results or caches to registries, local storage, and cloud backends.

SetupDocsCommunityMaturity

Verdict Our BuildKit checkout built in 100 seconds, but 13 of 104 package tests failed in the worker area on fresh unprivileged Debian, so a green source build does not finish the evaluation. Run standalone BuildKit when remote workers, LLB frontends, or controlled cache exports justify operating a privileged builder service.

Full review →
219

checkstyle

checkstyle/checkstyle · ★ 9.6k2/7d · Java

Checkstyle checks Java source code against a chosen coding standard or set of best practices. It catches structural and style violations during development or CI, so teams can enforce consistent rules without relying on manual review alone.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 102 seconds, while the full test command ended with 20,607 tests passed and 12 collection/setup errors. Checkstyle is a mature, actively maintained choice when a Java team needs policy enforcement rather than automatic formatting or defect discovery.

Full review →
220

vtracer

visioncortex/vtracer · ★ 7.0k96/7d · Rust

VTracer converts bitmap images such as JPEG and PNG files into SVG vector graphics. It handles colour artwork and photographs as well as black-and-white line art, with desktop, command-line, Rust, Python, and Node interfaces built around the same tracing engine.

SetupDocsCommunityMaturity

Verdict Our VTracer run installed 83 packages in 9 seconds, built in 38 seconds, and passed all 172 tests in 51 seconds, making it the easiest repo here to trust for a trial. Use it for automated SVG tracing when colour handling, fixed palettes, or several language bindings matter.

Full review →
221

zizmor

zizmorcore/zizmor · ★ 6.5k55/7d · Rust

zizmor is a security-focused static analyzer for GitHub Actions, Dependabot configuration, and experimental pre-commit inputs. It reads CI files without executing them and reports risky permissions, expression injection, unpinned actions, credential exposure, confusing refs, and other mistakes that can turn a pull request into code execution.

SetupDocsCommunityMaturity

Verdict Our zizmor install took 352 seconds, while the build needed only 5 seconds and pip-audit found 0 known vulnerabilities in the installed Python packages. Use a packaged binary instead of building this source checkout unless you are contributing, then run the regular persona on every workflow change.

Full review →
222

destructive_command_guard

Dicklesworthstone/destructive_command_guard · ★ 6.0k47/7d · Rust

Destructive Command Guard, or dcg, is a Rust hook that inspects shell commands proposed by coding agents and blocks commands matching destructive rules before execution. It solves a narrow problem: an agent can move quickly enough to erase uncommitted work, cloud resources, or database data before a person catches the command.

SetupDocsCommunityMaturity

Verdict Our dcg run built successfully and passed 5,311 of 5,313 tests, but the 399-second suite still exited with 2 failures, so it is useful insurance rather than a safety boundary. Install it if agents can touch valuable working trees, then enable only the packs you can test against real commands.

Full review →
223

ESP32-Bit-Pirate

geo-tp/ESP32-Bit-Pirate · ★ 5.8k135/7d · C++

ESP32 Bit Pirate is open-source firmware that transforms a cheap ESP32-S3 board into a powerful hardware hacking multi-tool. It lets you interact with, debug, and analyze a massive range of digital and radio protocols, from common ones like I2C and SPI to Wi-Fi, Bluetooth, and Sub-GHz radio, all through a simple command-line interface in your browser or terminal.

SetupDocsCommunityMaturity

Verdict ESP32 Bit Pirate is an essential tool for any modern electronics workbench. It brilliantly resurrects the spirit of the original Bus Pirate, using the power and connectivity of the ESP32-S3 to create a tool that is vastly more capable and accessible.

Full review →
224

rayhunter

EFForg/rayhunter · ★ 5.8k29/7d · Rust

Rayhunter turns certain mobile hotspots and phones into monitors for suspicious cellular behavior associated with IMSI catchers. It records modem diagnostics, applies documented heuristics, and shows warnings through a local web interface, giving researchers evidence to inspect rather than a promise that every cell-site simulator will be found.

SetupDocsCommunityMaturity

Verdict Our Rayhunter build stopped after 147 seconds because embedded web assets were absent, and the test command hit the same compile failure in 16 seconds. Use the signed release installer on a recommended device if you need a practical field monitor, then treat every alert as evidence to investigate rather than a verdict.

Full review →
225

LACT

ilya-zlobintsev/LACT · ★ 5.6k40/7d · Rust

LACT is a Linux application for monitoring and changing AMD, Nvidia, and Intel GPU settings. Its graphical client talks to a system service that can manage clocks, power limits, fan curves, profiles, sensors, and historical data without keeping a desktop session open.

SetupDocsCommunityMaturity

Verdict Our LACT build stopped in 42 seconds because GTK 4.8.3 did not meet the required 4.14 floor, and the test command hit the same check before running tests. Use the packaged release when LACT's hardware matrix covers your card and you want Linux GPU controls that survive beyond one graphical session.

Full review →
226

avoid-ai-writing

conorbronsdon/avoid-ai-writing · ★ 4.4k174/7d · JavaScript

avoid-ai-writing is an Agent Skill and JavaScript detector for finding repetitive, inflated, or chatbot-like prose. It can audit text, rewrite it, or edit prose files in place, with profiles for different voices and writing contexts.

SetupDocsCommunityMaturity

Verdict Our run installed 0 packages, passed its tests in 10 seconds, and found 0 known vulnerabilities, so trying avoid-ai-writing costs almost nothing. Use the skill as a sharp editing checklist and the detector as a prompt for review.

Full review →
227

openscreen

getopenscreen/openscreen · ★ 2.9k345/7d · TypeScript

OpenScreen is a desktop recorder and editor for turning screen captures into product demos, tutorials, and short walkthroughs. It records the screen, microphone, system audio, and webcam, then adds cursor treatment, zooms, captions, backgrounds, annotations, and MP4 or GIF export on Windows, macOS, and Linux.

SetupDocsCommunityMaturity

Verdict Our OpenScreen run passed 2,687 of 2,688 tests, but its 46-second build failed and npm audit found 38 known vulnerabilities. Try the signed release if you want a free desktop demo editor with unusually deep capture and automation options, then run a disposable recording through your exact hardware before important work.

Full review →
228

git-ai

git-ai-project/git-ai · ★ 2.8k208/7d · Rust

Git AI is a local Git extension that records which lines a supported coding agent wrote, plus the agent, model, session, and prompt behind them. It provides AI-aware blame and commit statistics without trying to detect authorship from code after the fact.

SetupDocsCommunityMaturity

Verdict Our Git AI run passed 2,324 of 2,325 tests after a 293-second build, with the lone failure tied to an unsupported `git merge-tree` option in our sandbox. That is strong enough for an individual trial and cautious enough to demand workflow-specific checks before teams treat attribution as audit evidence.

Full review →
229

workmux

raine/workmux · ★ 2.6k246/7d · Rust

Workmux pairs each Git worktree with a tmux window or another supported terminal multiplexer, giving every branch its own shell, editor, server, and coding agent. It creates the workspace, runs setup commands, tracks agent status, and can merge and remove the branch when the task is finished.

SetupDocsCommunityMaturity

Verdict Our run of Workmux built successfully in 100 seconds and all 3,144 tests passed in 40 seconds, a strong outcome for a tool that can create, merge, and delete working environments. Use it when tmux and Git worktrees are already habits and parallel agents are becoming hard to track.

Full review →
230

gh-stack

github/gh-stack · ★ 1.5k17/7d · Go

GitHub Stacked PRs is a GitHub CLI extension for turning a large change into an ordered chain of smaller pull requests. It tracks the branches locally, keeps their bases in order, and can submit, sync, restructure, navigate, or merge the stack from the terminal.

SetupDocsCommunityMaturity

Verdict Our gh-stack run installed 203 packages and passed all 15 tests, so the extension is easy to trial but its v0.1.0 behavior still needs supervision. Use it when your team already understands rebases and wants GitHub-native stacked pull requests from the CLI.

Full review →
231

breakscale

xevrion/breakscale · ★ 1.2k · TypeScript

Breakscale is a browser-based simulator for learning how distributed systems behave as traffic rises or components fail. You draw a topology, change the load, and watch queues, latency, retries, and goodput move through a deterministic discrete-event model.

SetupDocsCommunityMaturity

Verdict Our Breakscale run built in 6 seconds and all 1,064 tests passed in 21 seconds, making it the cleanest lab result in this group. Use it to learn and discuss failure mechanics, especially queues and retry storms.

Full review →
232

usage

jdx/usage · ★ 1.0k8/7d · Rust

Usage is a specification, command-line program, and Rust framework for defining a CLI once and reusing that definition for parsing, help, shell completions, man pages, Markdown, and generated SDKs. It solves the drift that appears when a tool's parser, documentation, and completion scripts are maintained separately.

SetupDocsCommunityMaturity

Verdict Our npm run installed 274 packages in 35 seconds, then found no standard build or test target, so it verified the documentation toolchain's dependencies rather than the Rust product. Choose Usage when one CLI contract must feed help, completions, docs, SDKs, and agents.

Full review →
233

skills

mattpocock/skills · ★ 263k5.0k/7d · Shell

Matt Pocock's Skills is a collection of plain-text instructions that gives coding agents repeatable workflows for specification, debugging, testing, review, and project handoffs. You can install the full set as a managed Claude Code plugin or copy selected skills into a project for use with Codex and other agents.

SetupDocsCommunityMaturity

Verdict Use this collection if you already trust coding agents with real repository work and want their process written down where you can inspect and change it. The individual skills are unusually specific about feedback loops and approval points, but they remain prompts rather than enforcement.

Full review →
234

linux

torvalds/linux · ★ 249k1.5k/7d · C

This is the mainline source tree for the Linux kernel, the code that manages processors, memory, devices, filesystems, networking, and the boundary used by programs on Linux systems. It is source code for kernel development and integration rather than a packaged Linux distribution.

SetupDocsCommunityMaturity

Verdict Our run built only Linux's documentation helper in 7 seconds and did not compile or test the kernel, so it says nothing about kernel build success at commit 893e117. Use this tree when you need mainline kernel source, upstream review, or the exact implementation behind Linux behavior.

Full review →
235

HelloGitHub

521xueweihan/HelloGitHub · ★ 177k1.2k/7d · Python

HelloGitHub is primarily a Chinese-language monthly publication that introduces interesting, beginner-friendly open-source projects, books, practical projects, and enterprise projects. English and Japanese README translations exist, but the main issue archive shown here is Chinese, so English-only readers should expect a language barrier.

SetupDocsCommunityMaturity

Verdict Our build completed in 7 seconds after a 17-second install, but there was no test target to run. Use HelloGitHub as a lively Chinese-language discovery publication, not as software your team needs to adopt.

Full review →
236

rust

rust-lang/rust · ★ 119k901/7d · Rust

rust-lang/rust is the main source repository for the Rust compiler, standard library, and official language documentation. Application developers normally install released Rust toolchains through rustup; this checkout is for compiler work, standard-library changes, platform ports, documentation, and source builds.

SetupDocsCommunityMaturity

Verdict Our harness installed 138 Yarn packages in 33 seconds but compiled no Rust code and ran no Rust tests, so it says nothing about rustc health. Use the released Rust toolchain for application development; clone this 3,489.9 MB source tree only when the compiler, standard library, documentation, or a supported platform is the work itself.

Full review →
237

rtk

rtk-ai/rtk · ★ 81k852/7d · Rust

RTK is a Rust command-line proxy that shortens output from Git, test runners, linters, package managers, and other developer commands before an AI coding agent reads it. It solves a narrow problem: noisy shell output consumes model context, while the developer still needs failures, counts, and useful file details.

SetupDocsCommunityMaturity

Verdict Our RTK run installed 160 packages and passed all 2,704 tests, so it earns a trial for developers whose coding agents consume long shell output. Begin with interactive, read-heavy commands and keep raw recovery enabled.

Full review →
238

ghidra

NationalSecurityAgency/ghidra · ★ 77k2.1k/7d · Java

Ghidra is a desktop and headless reverse-engineering toolkit for taking apart compiled programs. It combines disassembly, decompilation, graphing, scripting, and extension APIs so analysts can understand binaries without source code.

SetupDocsCommunityMaturity

Verdict Our Ghidra install step took 29 seconds, but both the 12-second build and 10-second test run stopped because the local Maven repository was missing. Use the official archive if you need the analyst tool; use the source tree when you are prepared to follow its separate dependency-fetch and development setup.

Full review →
239

moby

moby/moby · ★ 72k29/7d · Go

Moby is the open-source engine code and component set used upstream by Docker Engine and other container systems. It gives engine builders APIs and swappable pieces for images, containers, networking, storage, and orchestration, rather than a polished product for ordinary container users.

SetupDocsCommunityMaturity

Verdict Our Moby build took 188 seconds, while tests hit the 900-second cap with 19 passing and 4 failing out of 23 recorded. That cost makes sense for engineers changing Docker Engine or building a related container platform.

Full review →
240

archify

tt-a1i/archify · ★ 64k8.9k/7d · HTML

Archify is an agent skill that turns a repository or system description into a polished, interactive architecture map. It is meant to replace hand-maintained diagrams with typed, checked artifacts that can be searched, presented, compared, and exported.

SetupDocsCommunityMaturity

Verdict Our run installed Archify in 22 seconds, but 2 of 710 tests failed, so it is promising rather than friction-free. Use it if interactive, source-grounded maps would improve design reviews or onboarding and you can validate the generated artifact in your own environment.

Full review →
241

drawio-desktop

jgraph/drawio-desktop · ★ 63k123/7d · JavaScript

draw.io Desktop is the official Electron wrapper for the draw.io diagram editor, packaged for Windows, macOS, and Linux. It lets people make flowcharts, network diagrams, UML, and other visual documents in local files without an account or a cloud service.

SetupDocsCommunityMaturity

Verdict Our draw.io Desktop run installed 323 packages in 11 seconds, then passed all 158 tests in 5 seconds with 0 audit findings. That is strong evidence for trying the official binary when you need an offline, account-free diagram editor.

Full review →
242

Magisk

topjohnwu/Magisk · ★ 63k138/7d · Kotlin

Magisk is a Kotlin-led, open-source toolkit for customizing rooted Android devices, with English documentation available. It gives apps controlled root access, applies modifications without directly rewriting read-only partitions, handles boot images, and can load code into app processes.

SetupDocsCommunityMaturity

Verdict Our install attempt failed after 11 seconds because stable Cargo could not accept a nightly-only feature. Magisk is still the strongest default here for experienced Android modders who need its particular combination of root management, systemless modules, boot-image tooling, and Zygisk.

Full review →
243

marktext

marktext/marktext · ★ 61k · TypeScript

MarkText is a free desktop Markdown editor that shows formatted text as you write while keeping the document as a normal local `.md` file. It runs on Linux, macOS, and Windows, with source, focus, and typewriter modes for people who want a calmer alternative to a code editor.

SetupDocsCommunityMaturity

Verdict Our MarkText install took 138 seconds and 1,735 MB, and 815 of 818 tests passed before 3 PDF tests failed, so the app is easier to recommend as a downloaded editor than as a low-friction source dependency. Use it for focused work on local Markdown files when inline formatting matters more than notebooks, sync, or collaboration.

Full review →
244

ghostty

ghostty-org/ghostty · ★ 61k298/7d · Zig

Ghostty is a terminal emulator for macOS and Linux that combines a platform-native interface with GPU-rendered terminal sessions. The same codebase also provides `libghostty-vt`, a C-compatible library for applications that need terminal parsing and state without building that machinery themselves.

SetupDocsCommunityMaturity

Verdict Our Ghostty install failed after 7 seconds because CMake could not find `zig`, so source development starts with matching the documented toolchain before any build or test can run. Use the packaged app if you want a polished native terminal on macOS or Linux; the active repository and broad protocol support make it worth a serious trial.

Full review →
245

ImHex

WerWolv/ImHex · ★ 55k55/7d · C++

ImHex is a graphical hex editor and binary-analysis workbench for reverse engineers, programmers, and file-format researchers. It combines byte editing with a C-like pattern language, disassembly, diffing, hashing, YARA scans, visualizers, remote data sources, and plugins.

SetupDocsCommunityMaturity

Verdict Our ImHex source install stopped after 4 seconds because CMake could not find `dbus-1`, so most users should start with a packaged build rather than compile it cold. Choose ImHex when its Pattern Language and many data sources will replace several smaller binary tools in your workflow.

Full review →
246

kotlin

JetBrains/kotlin · ★ 53k35/7d · Kotlin

Kotlin is JetBrains' concise, multiplatform programming language and the main repository for its compiler, standard libraries, and supporting build infrastructure. It helps teams share code across JVM, Android, iOS, desktop, web, and native targets while still allowing platform-specific work where needed.

SetupDocsCommunityMaturity

Verdict Our 19-second install succeeded with 24 npm packages, but there was no build or test target in the measured WASM browser workspace. Use Kotlin when Android, JVM interoperability, or shared multiplatform code is a core requirement and your team can accept a large, sophisticated toolchain.

Full review →
247

cypress

cypress-io/cypress · ★ 51k13/7d · TypeScript

Cypress is a browser testing system for end-to-end and component tests, with an interactive runner for watching commands, page state, requests, and failures. It gives web teams one JavaScript or TypeScript workflow for local debugging and automated browser checks.

SetupDocsCommunityMaturity

Verdict Our Cypress run consumed 3,249 MB and 525 seconds for installation, built in 211 seconds, then failed with 5 of 7 parsed Mocha tests failing after 106 seconds. The product remains a strong choice for teams that value its interactive runner and browser-centered debugging enough to absorb a heavy toolchain.

Full review →
248

open-source-mac-os-apps

serhii-londar/open-source-mac-os-apps · ★ 50k96/7d

Awesome macOS Open Source Applications is a curated directory of free, open-source Mac software, not an application you install or run. It organizes 703 native and cross-platform apps into 49 categories, helping Mac users discover alternatives and developers find projects to contribute to.

SetupDocsCommunityMaturity

Verdict Our run produced no install, build, or test result because commit 04deaf8 exposes no supported software ecosystem and includes no Dockerfile. Use this repository as a strong discovery index, especially if open source and macOS are non-negotiable, but do not mistake its 703 entries for 703 endorsements.

Full review →
249

bruno

usebruno/bruno · ★ 47k121/7d · JavaScript

Bruno is an open-source desktop client for sending API requests, organizing collections, and running those collections from a CLI. It keeps collection data in plain-text Bru files on your computer, so teams can review API changes through Git instead of depending on a hosted workspace.

SetupDocsCommunityMaturity

Verdict Our Bruno source install took 97 seconds, pulled 3,430 packages, and left 1,956 MB on disk, so developers should use a packaged release unless they intend to contribute. For teams that want API collections reviewed in Git and kept off a vendor cloud, Bruno is an easy recommendation.

Full review →
250

helix

helix-editor/helix · ★ 46k94/7d · Rust

Helix is a terminal code editor with modal commands, multiple selections, built-in language-server support, and tree-sitter parsing. It uses Kakoune's selection-first editing model, so it solves the usual terminal-editor setup problem without copying Vim's command grammar.

SetupDocsCommunityMaturity

Verdict Our Helix run installed 327 packages in 36 seconds, but neither the build nor tests finished within 900 seconds, so source-level evaluation needs more time than a quick editor trial. Use a packaged binary if you want a terminal-first editor with language-server support and can accept selection-first commands.

Full review →
251

Files

files-community/Files · ★ 45k487/7d · C#

Files is an open-source file manager made specifically for Windows. It replaces the single-window feel of File Explorer with tabs, dual panes, tags, custom layouts, cloud and FTP access, and integrations such as Git while keeping a Windows-style interface.

SetupDocsCommunityMaturity

Verdict We could not run Files because our sandbox has no supported C# path and the repository provides no Dockerfile, so this review cannot confirm its build or tests. The released app is a serious File Explorer alternative for Windows users who value tabs, dual panes, tags, and Git-aware folders.

Full review →
252

grpc

grpc/grpc · ★ 45k7/7d · C++

gRPC is a remote procedure call framework for connecting services through typed contracts instead of hand-written HTTP payloads. This repository holds the shared C++ core plus the C++, Python, Ruby, PHP, Objective-C, and older core-based C# implementations; several other language runtimes live in separate repositories.

SetupDocsCommunityMaturity

Verdict Our gRPC source install ran for 900 seconds without finishing, so adopting the protocol is far easier than becoming a contributor to this 632.7 MB repository. Use the published runtime for your language when typed service contracts and streaming RPC justify generated code.

Full review →
253

reactjs-interview-questions

sudheerj/reactjs-interview-questions · ★ 45k25/7d · JavaScript

ReactJS Interview Questions is an English reference that collects short answers, code samples, and practice material for React interviews. It covers core React, Router, testing, Redux, React Native, React 18 and 19, with older answers kept in a separate section.

SetupDocsCommunityMaturity

Verdict Our run installed 22 packages in 15 seconds but found no root build or test target, so this is a reading list first and a runnable project collection second. Use it to make a study checklist or start an interview discussion, especially when React 19 and Redux questions matter.

Full review →
254

sentry

getsentry/sentry · ★ 45k28/7d · Python

Sentry is an error tracking and application monitoring platform that connects crashes, traces, logs, replays, and production context in one debugging workspace. It helps developers move from a reported failure to the code and request that caused it.

SetupDocsCommunityMaturity

Verdict Our Sentry build succeeded in 93 seconds, but its test run passed the 900-second cap, so this is mature product infrastructure with a contributor loop sized for dedicated teams. Choose it when errors need to connect to traces, replays, logs, and ownership workflows across many applications.

Full review →
255

sniffnet

GyulyVGC/sniffnet · ★ 41k59/7d · Rust

Sniffnet is a desktop application for watching the network traffic entering and leaving one computer. It turns packet captures into searchable connections, charts, host details, alerts, and PCAP reports without asking the user to learn Wireshark first.

SetupDocsCommunityMaturity

Verdict Our Sniffnet build stopped after 249 seconds because the linker could not find `libpcap`, so source users should install the documented native capture packages before judging the Rust code. The released app is a good choice for people who want local traffic explained in a friendly desktop interface, especially when Wireshark feels excessive.

Full review →
256

styled-components

styled-components/styled-components · ★ 41k · TypeScript

styled-components lets React developers write CSS inside JavaScript or TypeScript components, generating scoped class names and styles from props. It covers browser React, server rendering, Server Components, and React Native through related APIs.

SetupDocsCommunityMaturity

Verdict Our styled-components checkout passed all 1,465 tests after a 90-second build, but its contributor install occupied 1,218 MB. Use it when React components genuinely need prop-aware styles, themes, and shared web/native conventions.

Full review →
257

insomnia

Kong/insomnia · ★ 40k18/7d · TypeScript

Insomnia is a desktop API client for sending and debugging REST, GraphQL, WebSocket, SSE, gRPC, and other HTTP-compatible requests. It also handles OpenAPI design, automated request tests, mocks, and command-line checks, so an API team can keep exploration and repeatable validation in one project.

SetupDocsCommunityMaturity

Verdict Our Insomnia run installed 1,750 packages, passed all 18 available tests, and still reported 39 known dependency vulnerabilities, so it is easy to try but costly to adopt casually as source. Use the packaged desktop app if you need one client for REST, GraphQL, gRPC, WebSockets, and OpenAPI work, especially when local or Git-backed storage matters.

Full review →
258

googletest

google/googletest · ★ 40k36/7d · C++

GoogleTest is a C++ unit-testing framework, bundled with GoogleMock for writing mocks. It gives C++ teams assertions, automatic test discovery, parameterized tests, death tests, and a familiar xUnit structure without making them design a test runner.

SetupDocsCommunityMaturity

Verdict Our GoogleTest install stopped after 7 seconds because CMake could not find Abseil, so commit 36ba75f is a poor blind clone-and-build choice in fresh Debian. The framework itself remains a sensible default for C++ teams that need mature assertions and mocking, especially when their package manager or build system already supplies its dependencies.

Full review →
259

aseprite

aseprite/aseprite · ★ 39k151/7d · C++

Aseprite is a desktop editor for drawing pixel art and frame-based sprite animation, with layers, palettes, onion skinning, tiled drawing, sprite-sheet export, Lua scripting, and a command-line interface. Its source is public, but the project EULA restricts modification and redistribution, so it is source-available rather than open source under a standard OSI license.

SetupDocsCommunityMaturity

Verdict Our Aseprite checkout exposed only a `third_party/benchmark` Python build, which took 5 seconds and then found 0 tests, so it provides no evidence that the C++ editor builds cleanly. The editor itself is an easy recommendation for focused pixel-art production if its timeline, export, scripting, and polished native workflow match the team.

Full review →
260

devdocs

freeCodeCamp/devdocs · ★ 39k26/7d · Ruby

DevDocs puts reference manuals for many programming languages and tools into one searchable web interface. It solves the tab-hopping problem with fast lookup, keyboard navigation, and offline copies, while deliberately leaving tutorials and broad web search to other tools.

SetupDocsCommunityMaturity

Verdict Our 2026-09-02 sandbox did not run DevDocs because the lab lacked a supported Ruby ecosystem, so we cannot vouch for its install, build, or tests. Use the hosted service if you want fast reference lookup with offline support, and self-host only if you can own the Ruby scraper and two-part update process.

Full review →
261

metasploit-framework

rapid7/metasploit-framework · ★ 39k38/7d · Ruby

Metasploit Framework is an open-source penetration-testing console and module system for checking vulnerabilities, running authorized exploits, creating payloads, and working with resulting sessions. It puts many security-testing workflows behind one searchable interface instead of making an operator assemble separate scripts for each target.

SetupDocsCommunityMaturity

Verdict Our Metasploit checkout used 201.6 MB before dependencies, and its nested npm project added 330 MB with 3 known vulnerabilities, so adopting the source tree means owning a large, mixed dependency surface. Use it when an authorized team needs a shared console and module system across many kinds of security tests.

Full review →
262

cli

httpie/cli · ★ 39k28/7d · Python

HTTPie CLI is a terminal HTTP client built for people who inspect and send API requests by hand. Its `http` and `https` commands turn headers, JSON fields, forms, files, authentication, sessions, downloads, and formatted responses into readable command lines.

SetupDocsCommunityMaturity

Verdict Our HTTPie CLI run passed 1,016 of 1,019 tests, but the same environment reported 16 known vulnerabilities, so convenience does not erase the need to review its installed dependency set. HTTPie remains one of the clearest tools for exploratory API calls and readable terminal output.

Full review →
263

leetcode

doocs/leetcode · ★ 37k39/7d · Java

doocs/leetcode is a Chinese-first library of LeetCode and programming-interview solutions, and an English README is available. It helps developers study algorithms, compare implementations across several languages, and find worked answers by problem or technique.

SetupDocsCommunityMaturity

Verdict Our pnpm install completed in 13 seconds, but there was no build or test target to validate the 610,402-line checkout. Use doocs/leetcode as a broad, actively maintained study reference, especially if you read Chinese and like comparing languages.

Full review →
264

wails

wailsapp/wails · ★ 36k35/7d · Go

Wails turns a Go backend and a web frontend into a desktop application that uses the operating system's webview. It gives Go developers an Electron-like workflow without shipping a separate browser engine inside every app.

SetupDocsCommunityMaturity

Verdict Our Wails v2 build finished in 33 seconds, but 4 of 28 tested packages failed in the 25-second suite, so adopters should reproduce the full cross-platform matrix before shipping. Wails is a good fit when the backend is already Go and a web UI is the fastest route to a desktop product.

Full review →
265

awesome-react-native

jondot/awesome-react-native · ★ 36k2/7d

Awesome React Native is an English-language, editor-curated directory of libraries, tools, apps, and learning material for React Native developers. It helps you assemble a shortlist by grouping current options around jobs such as navigation, storage, testing, and on-device AI, but it does not compare them or guarantee compatibility with your app.

SetupDocsCommunityMaturity

Verdict We could not run commit 6733606 because this 28,684-byte README has no supported software ecosystem or Dockerfile, so Awesome React Native should be used as a reading list, never as tested tooling. Its August 26, 2026 rebuild makes it a useful first pass for developers who have not tracked the ecosystem recently.

Full review →
266

netty

netty/netty · ★ 35k11/7d · Java

Netty is a Java framework for building network servers and clients that need to handle many connections without assigning a thread to every request. It supplies the low-level networking foundation so application teams can focus on protocols and business behavior instead of socket plumbing.

SetupDocsCommunityMaturity

Verdict Our build failed after 112 seconds, while 6,941 tests passed before setup errors stopped a clean test result. Netty remains a serious choice when a Java system genuinely needs a low-level asynchronous networking foundation and the team can own its operational complexity.

Full review →
267

swc

swc-project/swc · ★ 34k3/7d · Rust

SWC is a Rust compiler and transformation platform for JavaScript and TypeScript, exposed to both Rust and JavaScript projects. It parses, transpiles, minifies, and rewrites code quickly enough to sit inside build tools, while its packages and configuration cover common module and syntax targets.

SetupDocsCommunityMaturity

Verdict Our SWC install finished in 53 seconds, but the build failed after 16 seconds and the test run hit 900 seconds while Cargo-backed plugin work remained, so contributing requires the full Rust and WebAssembly environment. Use the published packages when a supported framework needs fast transpilation; keep `tsc` for type checking and regression-test emitted code.

Full review →
268

FreeCAD

FreeCAD/FreeCAD · ★ 34k177/7d · C++

FreeCAD is a desktop application for parametric 3D modeling, where dimensions and earlier construction steps remain editable instead of becoming fixed geometry. It is aimed at product design, mechanical engineering, architecture, education, and hobby work on Windows, macOS, and Linux.

SetupDocsCommunityMaturity

Verdict Our 34-second Pivy install and 8-second build succeeded, but all 4 test modules stopped at setup because `pivy.coin` was missing. FreeCAD is still the open-source CAD project to try first when you need serious parametric 3D work across desktop platforms, especially if Python scripting matters.

Full review →
269

Proton

ValveSoftware/Proton · ★ 33k52/7d · C++

Proton lets Steam run Windows-only games on Linux by using Wine underneath. For most players, it solves compatibility inside Steam without asking them to build, patch, or configure the project themselves.

SetupDocsCommunityMaturity

Verdict Our build failed with exit 101 after 20 seconds because gstreamer-1.0 was missing. That makes source compilation a poor first route for ordinary players, especially when Valve explicitly says to use the Proton version supplied by Steam.

Full review →
270

jj

jj-vcs/jj · ★ 32k106/7d · Rust

Jujutsu is a command-line version control system that stores ordinary Git commits while replacing Git's staging, branch, conflict, and history-editing workflow. It is for developers who want changes to stay easy to rewrite and repository operations to remain undoable without giving up Git remotes.

SetupDocsCommunityMaturity

Verdict Our Jujutsu run installed 33 packages in 11 seconds and built in 1 second, but its tests were skipped because the harness found no test script or target. Experienced Git users who constantly rewrite commit stacks should trial it in a colocated repository, where the operation log and automatic rebasing can repay the learning cost.

Full review →
271

Avalonia

AvaloniaUI/Avalonia · ★ 32k51/7d · C#

Avalonia is a C# and XAML framework for building desktop, mobile, embedded, and WebAssembly interfaces from a shared codebase. It gives .NET teams a WPF-like development model across Windows, macOS, Linux, iOS, Android, and the browser, while leaving a true WPF compatibility route to the separate commercial XPF product.

SetupDocsCommunityMaturity

Verdict Our sandbox did not run Avalonia at commit 38402fc because C# was outside the harness and the repository had no Dockerfile. That leaves setup ease unverified, but the documented NuGet path and active 12.1.1 release make Avalonia a serious choice for a .NET team willing to test every target OS itself.

Full review →
272

iced

iced-rs/iced · ★ 31k32/7d · Rust

Iced is a Rust library for building graphical applications on Windows, macOS, Linux, and the web from one state-and-message model. It handles widgets, layout, events, windows, and rendering so a Rust developer does not have to assemble those pieces from lower-level graphics and windowing crates.

SetupDocsCommunityMaturity

Verdict Our Iced build took 149 seconds and all 70 tests passed in 55 seconds, so the measured checkout is healthy enough for a serious prototype. Use it for a Rust-first desktop UI when the Elm-style model and custom rendering are assets, and budget for platform testing.

Full review →
273

niri

niri-wm/niri · ★ 28k · Rust

Niri controls a Wayland desktop by arranging windows in columns on a horizontal strip that you scroll. New windows appear to the right without squeezing the ones already open, and every monitor keeps its own strip and workspaces. It supplies the window-management core while leaving panels, notifications, login prompts, and other desktop pieces to separate programs.

SetupDocsCommunityMaturity

Verdict Our 30-second niri install added 408 packages, but both build and test commands exited 101 because pkg-config could not find `libpipewire-0.3`. Try the distribution package if per-monitor scrolling columns solve a problem you actually have, especially when ordinary tilers keep rearranging useful windows.

Full review →
274

cmux

manaflow-ai/cmux · ★ 27k224/7d · Swift

cmux is a macOS terminal built with Swift and AppKit on top of libghostty. It adds vertical workspaces, split terminals, agent notifications, an embedded browser, SSH workspaces, and a socket API for developers running several coding agents at once.

SetupDocsCommunityMaturity

Verdict Our cmux checkout installed 457 packages and used 510 MB, but it exposed no build or test target in the measured Bun project. The shipped macOS app is worth trying for developers supervising several terminal agents because notifications, workspaces, and browser control address that exact job.

Full review →
275

UniGetUI

Devolutions/UniGetUI · ★ 26k78/7d · C#

UniGetUI is a desktop interface that lets you find, install, update, and remove software managed by several package managers from one place. It is primarily for Windows 10 and 11, while also offering macOS and Linux builds for people who would rather use a graphical app than remember each manager's commands.

SetupDocsCommunityMaturity

Verdict Our run produced no install, build, or test result because the Debian sandbox did not support C# and the repository offered no Dockerfile. For Windows 10 or 11 users juggling several package managers, UniGetUI is still an unusually practical front end, and the active September 2026 development makes it worth trying from an official channel.

Full review →
276

bazel

bazelbuild/bazel · ★ 26k35/7d · Java

Bazel is a build and test system for repositories that span multiple languages, platforms, or very large dependency graphs. It avoids repeating work through dependency analysis, local or remote caching, and parallel execution, while letting teams define build rules in Starlark.

SetupDocsCommunityMaturity

Verdict Our 220.3 MB Bazel checkout exposed no test target to the generic harness, and the 6-second step it built was not the main Java build system. Bazel is worth the adoption cost when one dependency graph, remote caching, and several languages can repay dedicated build-engineering work.

Full review →
277

v8

v8/v8 · ★ 25k12/7d · C++

V8 is Google's open-source JavaScript engine, written in C++ and built to implement the ECMAScript language specification. It powers JavaScript in Chromium and can also be embedded in standalone C++ applications that need a high-performance language runtime.

SetupDocsCommunityMaturity

Verdict Our targeted build completed in 2 seconds, but pytest exited 5 after finding 0 tests, so this checkout did not give us end-to-end validation of V8 itself. Use V8 when Chromium-grade ECMAScript compatibility, active development, and serious C++ embedding capability justify its heavyweight workflow.

Full review →
278

compound-engineering-plugin

EveryInc/compound-engineering-plugin · ★ 25k126/7d · TypeScript

Compound Engineering is a set of 33 reusable instructions that gives AI coding agents a fixed loop for brainstorming, planning, implementation, review, and recording lessons. It is for teams that want an agent to follow a repeatable engineering process instead of improvising each task from a short prompt.

SetupDocsCommunityMaturity

Verdict Our test command ran for 455 seconds and reported 177 failures at commit 26bf5b1, so Compound Engineering should earn trust in a trial repository before its autonomous workflow receives permission to commit or push. Use it if your team wants a strongly prescribed loop and will keep its plans and solution notes in the codebase.

Full review →
279

radare2

radareorg/radare2 · ★ 25k40/7d · C

radare2 is a command-line reverse-engineering framework for inspecting, debugging, disassembling, emulating, and modifying binaries. It gives security researchers a Unix-like set of small tools, libraries, plugins, and a scriptable interactive shell instead of centering the job on one graphical application.

SetupDocsCommunityMaturity

Verdict Our radare2 npm install ran for 415 seconds before a `wasm-pack` download timed out, so that packaging route is a poor first choice on restricted networks. The native tool remains one of the deepest options for reverse engineers who think in commands and scripts, with current releases and a huge format surface.

Full review →
280

fastfetch

fastfetch-cli/fastfetch · ★ 25k78/7d · C

Fastfetch is a C command-line program that reads local system details and prints them beside a logo in the terminal. It is a maintained replacement for Neofetch with JSONC configuration, machine-readable output, and support for Linux, macOS, Windows, Android, BSD variants, Haiku, and SunOS.

SetupDocsCommunityMaturity

Verdict Our fastfetch source install stopped after 10 seconds because CMake required C23 and the sandbox compiler did not provide it. Use a current prebuilt package if you want the quickest Neofetch replacement, or bring a documented C23 toolchain if you must compile it.

Full review →
281

loguru

Delgan/loguru · ★ 24k13/7d · Python

Loguru is a Python logging library built around one preconfigured logger and one method for adding output destinations. It makes colored console output, rotating files, structured records, exception capture, and contextual fields easier to set up than Python's standard logging package.

SetupDocsCommunityMaturity

Verdict Our Loguru sandbox installed 103 packages and passed its test command with 1,669 reported passes, but the package build exited after 8 seconds without a supplied error excerpt. Use it for Python applications where a single global logger and concise sink setup are worth the migration work.

Full review →
282

mihon

mihonapp/mihon · ★ 24k · Kotlin

Mihon is an Apache-2.0 Android reader for manga, webtoons, comics, and other image-based publications stored locally or supplied through extensions. It organizes a library, downloads chapters, schedules updates, and sends reading progress to 8 listed tracking services. Mihon ships no content or extension repository, so users must bring local files or choose third-party sources themselves.

SetupDocsCommunityMaturity

Verdict Our Mihon install step took 203 seconds, but both the 17-second build and 14-second test steps stopped because the Android SDK path was unset. For Android readers who can vet their own extension source, Mihon remains the sensible default: the reader, library, downloads, trackers, and backup format are well documented.

Full review →
283

jdk

openjdk/jdk · ★ 23k18/7d · Java

OpenJDK's JDK repository is the mainline source for the Java Development Kit, including the virtual machine, compiler, class libraries, and service tools. It is where contributors change Java itself; application developers who only need a working JDK are directed to prebuilt distributions.

SetupDocsCommunityMaturity

Verdict Our OpenJDK install stopped after 5 seconds with exit 126 before a build could begin, so this 694 MB source tree demands a prepared JDK development machine. Use it when you intend to change Java, maintain a port, or produce a downstream JDK.

Full review →
284

maui

dotnet/maui · ★ 23k · C#

.NET MAUI is Microsoft's C# and XAML framework for building native Android, iOS, macOS, and Windows applications from one shared project. It provides common UI controls and device APIs while still allowing platform-specific code when one operating system behaves differently.

SetupDocsCommunityMaturity

Verdict Our lab did not run `dotnet/maui`: the harness had no supported C# ecosystem and commit b96aa03 supplied no Dockerfile. MAUI is the default shortlist choice for a .NET team shipping native mobile and desktop apps, provided a Mac and platform-specific QA are budgeted.

Full review →
285

mocha

mochajs/mocha · ★ 23k · JavaScript

Mocha is a test framework for JavaScript projects running on Node.js or in a browser. It gives teams a familiar place to write and run checks so regressions are caught before software ships.

SetupDocsCommunityMaturity

Verdict Our build finished in 7 seconds, but the test command still exited 1 after 313 seconds despite reporting 374 passed and 0 failed. Mocha remains a credible default for teams that value a mature, familiar Node.js and browser test framework with visible maintenance.

Full review →
286

oxc

oxc-project/oxc · ★ 23k87/7d · Rust

Oxc is a Rust-based toolchain for JavaScript and TypeScript. It provides the Oxlint linter and Oxfmt formatter for application teams, plus parser, transformer, minifier, resolver, and code-generation pieces for developers building other tools.

SetupDocsCommunityMaturity

Verdict Our Oxc install took 31 seconds, but the build failed in 7 seconds and the test command failed in 5 seconds before running any tests, so the source checkout did not validate in a Node-only sandbox. Oxlint is still worth a side-by-side CI trial because released binaries avoid most monorepo setup and the project is intensely active.

Full review →
287

flow

facebook/flow · ★ 22k4/7d · Rust

Flow is a static type checker for JavaScript. It lets teams annotate selected files, catches type mistakes before runtime, and keeps checking changes through a background process.

SetupDocsCommunityMaturity

Verdict Our sandbox installed Flow's 126 Yarn packages in 20 seconds, but the checkout exposed no build or test target to run. Existing Flow teams should keep using the pinned checker and judge upgrades against their own code, especially around stricter releases.

Full review →
288

frida

frida/frida · ★ 22k64/7d · Meson

Frida lets developers and security researchers inspect and change a running program without rebuilding it. It solves the practical problem of seeing what software is doing at runtime, whether that software is on a desktop, mobile device, or another supported target.

SetupDocsCommunityMaturity

Verdict Our build completed in 10 seconds after a 42-second install, but the absence of a runnable test target keeps that clean result from being a full confidence signal. Use Frida when live, scriptable inspection is central to the job and you can tolerate a systems-heavy toolchain.

Full review →
289

gleam

gleam-lang/gleam · ★ 22k26/7d · Rust

Gleam is a statically typed programming language that compiles to Erlang or JavaScript. It aims to give application developers friendly compiler errors and type safety while keeping access to the Erlang and JavaScript ecosystems.

SetupDocsCommunityMaturity

Verdict Our Gleam build took 235 seconds and 12,129 of 12,137 tests passed, so the compiler is substantial and close to green in our sandbox, but contributors should reproduce the 8 failures before trusting a change. Gleam is a persuasive choice for teams that want static types on the Erlang VM and accept a younger ecosystem.

Full review →
290

lima

lima-vm/lima · ★ 22k61/7d · Go

Lima creates Linux virtual machines with automatic file sharing and port forwarding, mainly for running containers on macOS. It supplies the VM layer used by tools such as Colima, Rancher Desktop, Finch, and Podman Desktop, while remaining useful as a direct command-line tool.

SetupDocsCommunityMaturity

Verdict Our Lima build succeeded in 62 seconds, but 31 of 89 Go test packages failed in the same sandbox, so contributors should reproduce the suite before treating a checkout as ready. Lima is a strong VM foundation for macOS container tools and for developers who want control over images, drivers, mounts, and provisioning.

Full review →
291

graal

oracle/graal · ★ 22k10/7d · Java

Graal is the main source repository behind GraalVM, including its Java compiler, Native Image ahead-of-time compiler, Truffle language framework, WebAssembly engine, and several runtime tools. Most Java developers use a GraalVM distribution rather than building this repository; the source matters when you need to inspect or change those internals.

SetupDocsCommunityMaturity

Verdict Our Graal checkout built its detected target in 7 seconds with 0 audit findings, but no test target was available, so the run is a tooling check rather than proof of a complete GraalVM build. Most Java teams should download GraalVM and evaluate one application with Native Image instead of cloning this 3.5-million-line repository.

Full review →
292

obsidian-releases

obsidianmd/obsidian-releases · ★ 22k149/7d

This repository publishes Obsidian releases and maintains the directories that tell the app which community plugins and themes are available. It solves distribution and discovery for Obsidian's extension ecosystem, but it is not the Obsidian application source code and Obsidian itself is not open source.

SetupDocsCommunityMaturity

Verdict Our install completed in 11 seconds with 111 packages, but there was no build or test target because this is a distribution catalog, not the Obsidian source tree. Use it if you publish Obsidian plugins or themes, audit the community directory, or need official release metadata.

Full review →
293

OpenLogi

AprilNEA/OpenLogi · ★ 21k1.0k/7d · Rust

OpenLogi is a local desktop application and command-line tool for configuring Logitech mice, keyboards, lights, and webcams without a Logitech account or telemetry. It talks to supported hardware through HID++ or UVC and covers button remapping, DPI, SmartShift, profiles, lighting, and camera controls across macOS, Linux, and Windows.

SetupDocsCommunityMaturity

Verdict Our OpenLogi build took 82 seconds, and cargo test exited successfully after 52 seconds while discovering 0 tests, so package readiness is ahead of independent behavioral coverage in that command. Use it if local control and Linux support outweigh device-specific gaps, and verify every remap on your exact hardware.

Full review →
294

wasmer

wasmerio/wasmer · ★ 21k8/7d · Rust

Wasmer runs WebAssembly programs on desktops, servers, edge systems, and in browsers. It gives developers a portable execution layer with file, network, and environment access disabled unless they explicitly allow it, and it can also be embedded through SDKs for several languages.

SetupDocsCommunityMaturity

Verdict Our source build finished in 28 seconds, and all 4 tests that ran passed. Wasmer deserves a serious trial if you need a broad, embeddable WebAssembly platform with secure defaults, WASI and WASIX support, and distribution options beyond a bare runtime.

Full review →
295

intellij-community

JetBrains/intellij-community · ★ 21k27/7d · Java

IntelliJ Community is the open-source portion of JetBrains' IDE codebase and the foundation for IntelliJ Platform development. It is where contributors build IntelliJ IDEA Community, parts of PyCharm, platform services, language support, and IDE plugins against the real product source.

SetupDocsCommunityMaturity

Verdict Our 999.3 MB checkout contained 276,721 files, while the measured 12 passing tests covered only a small Go packer, so this repository demands more than its green lab result suggests. Use it when the work genuinely belongs inside the IntelliJ Platform or requires reading JetBrains' implementation.

Full review →
296

PEASS-ng

peass-ng/PEASS-ng · ★ 21k39/7d · C#

PEASS-ng is a suite of local privilege-escalation enumerators for Linux, Unix, macOS, and Windows. LinPEAS and WinPEAS inspect a host for permissions, credentials, services, software, and configuration mistakes that may give an authorized tester a route to higher privileges; optional parsers turn their colored output into JSON, HTML, or PDF.

SetupDocsCommunityMaturity

Verdict Our PEASS-ng run installed 35 packages in 18 seconds and built in 8 seconds, but there was no test target for the lab to run. Use the release artifacts for authorized host enumeration when an experienced tester will review both the commands and the findings.

Full review →
297

filament

google/filament · ★ 20k30/7d · C++

Filament is a C++ renderer that draws real-time 3D scenes with materials and lighting modeled on physical behavior. It gives Android, iOS, desktop, and web applications one rendering engine, plus tools for compiling materials and preparing scene assets.

SetupDocsCommunityMaturity

Verdict Our 1,537.2 MB Filament checkout installed its 45-package docs toolchain in 21 seconds and built those docs in 7 seconds, but pytest exited 5 after collecting 0 tests; we did not compile the C++ renderer. Use the release artifacts when a graphics team needs an embeddable 3D renderer across mobile and desktop and can own the material pipeline.

Full review →
298

blender

blender/blender · ★ 20k133/7d · C++

Blender is a free, open-source 3D creation suite covering modeling, rigging, animation, simulation, rendering, compositing, motion tracking, and video editing. This repository is the official GitHub mirror of its C and C++ source, intended for Blender development rather than the easiest route for artists who only want the application.

SetupDocsCommunityMaturity

Verdict Our Blender build failed after 8 seconds, and all 122 pytest failures shown in the 12-second run were blocked by the missing `bpy` module, so this checkout is unsuitable for a casual Python-style source install. Download the official binary if you want to create 3D work.

Full review →
299

vorssaint-utils

vorssaint/vorssaint-utils · ★ 19k2.1k/7d · Swift

Vorssaint is a free macOS menu bar app that combines per-app audio, window controls, system monitoring, clipboard history, screen capture, file utilities, and other everyday tools. It is meant to replace a pile of single-purpose Mac apps while letting you uninstall features you do not use.

SetupDocsCommunityMaturity

Verdict Vorssaint is compelling if five or six of its modules match utilities you already keep running. The feature installer and permission explanations make its breadth easier to control, but the fast release pace and current reports around the Command Bar, switcher, and Full Disk Access argue for a cautious rollout.

Full review →
300

chisel

jpillora/chisel · ★ 17k34/7d · Go

Chisel is a single-file networking tool that carries TCP or UDP tunnels over HTTP and protects the connection with SSH. It helps operators reach services through restrictive firewalls, expose a controlled route into a network, or create reverse forwards without deploying a larger VPN platform.

SetupDocsCommunityMaturity

Verdict Our build finished in 25 seconds, and all 18 tests passed. Chisel is an unusually capable answer when you specifically need SSH-secured TCP or UDP forwarding over HTTP in one portable executable.

Full review →
301

harper

Automattic/harper · ★ 15k183/7d · Rust

Harper is an English grammar and style checker that runs locally, including inside browsers, editors, Obsidian, a desktop app, and JavaScript applications. Its Rust core and WebAssembly build let it flag writing problems without sending the text to a remote grammar service.

SetupDocsCommunityMaturity

Verdict Our Harper install pulled 2,562 packages and used 1,629 MB, while the repository offered no root build or test target for us to verify. The finished extensions and local-only checking still make Harper worth trying for English writing where privacy comes first.

Full review →
302

portless

vercel-labs/portless · ★ 12k118/7d · TypeScript

Portless replaces changing local port numbers with stable names such as https://api.myapp.localhost. It runs a local reverse proxy, starts development servers on free ports, and keeps URLs predictable across monorepos and Git worktrees.

SetupDocsCommunityMaturity

Verdict Our Portless run installed 1,527 packages and used 1,061 MB, then all 931 executed tests passed in 83 seconds. That is a strong result for teams whose local stack has enough apps, worktrees, or OAuth callbacks to justify a machine-level proxy.

Full review →
303

gitoxide

GitoxideLabs/gitoxide · ★ 12k22/7d · Rust

Gitoxide is a Rust implementation of Git's storage, configuration, transport, revision, status, checkout, and repository plumbing. Application developers use the `gix` crate when they need to inspect or manipulate Git repositories without shelling out to the standard Git command.

SetupDocsCommunityMaturity

Verdict Our gitoxide build completed in 204 seconds and all 10 available cargo tests passed in 13 seconds after a 406-package install. Use `gix` when a Rust application needs deep, in-process Git access and the required operations are marked usable in the project's status documents.

Full review →
304

loop-engineering

cobusgreyling/loop-engineering · ★ 11k108/7d · JavaScript

Loop Engineering is a collection of patterns, starter files, and command-line tools for running coding agents on recurring jobs instead of prompting them by hand each time. It helps teams define state, budgets, safety limits, worktrees, and human approval points for agents such as Claude Code, Codex, Grok, and OpenCode.

SetupDocsCommunityMaturity

Verdict Our Loop Engineering checkout installed 8 packages in 3 seconds but offered no root build or test target, so it is easiest to assess as a pattern library rather than one verified application. Use it when recurring agent work needs written state, budgets, worktree isolation, and approval rules.

Full review →
305

kanata

jtroo/kanata · ★ 7.9k25/7d · Rust

Kanata is a cross-platform keyboard remapper for adding layers, tap-hold behavior, macros, Unicode output, and other programmable-keyboard ideas to ordinary keyboards. It runs on Linux, Windows, and macOS, translating input according to a text configuration file.

SetupDocsCommunityMaturity

Verdict Our Kanata build finished in 34 seconds, but 3 of 32 tests failed because their configuration files were not found. Kanata is a good fit for technical users who want deep keyboard layers on ordinary hardware and will test every mapping before relying on it.

Full review →
306

devenv

cachix/devenv · ★ 7.7k22/7d · Rust

devenv is a Nix-based tool for defining a project's packages, languages, services, processes, and checks in files that can be shared with the repository. It gives developers one command to enter the same working environment instead of maintaining separate setup instructions for each machine.

SetupDocsCommunityMaturity

Verdict Our devenv checkout installed 1,191 packages in 57 seconds, then its 219-second build and 46-second test run both failed on missing native tools. That makes devenv easier to recommend as an installed Nix product than as a Rust codebase you expect to compile in an ordinary Debian job.

Full review →
307

cloud-hypervisor

cloud-hypervisor/cloud-hypervisor · ★ 6.2k20/7d · Rust

Cloud Hypervisor is a Rust virtual machine monitor for modern 64-bit cloud guests on KVM or Microsoft's hypervisor. It favors virtio devices and limited legacy emulation, while supporting Linux and Windows guests, hotplug, snapshots, and machine-to-machine migration for cloud infrastructure builders.

SetupDocsCommunityMaturity

Verdict Our Cloud Hypervisor build took 176 seconds, and 295 of 323 tests passed before 28 io_uring-related failures ended the suite, so the code compiles but this checkout did not clear our test gate. Evaluate it for a modern 64-bit cloud fleet when minimal legacy emulation and migration features match your architecture.

Full review →
308

cli

docker/cli · ★ 6.1k18/7d · Go

Docker CLI is the Go client behind the `docker` command. It connects to a Docker daemon or context and gives users one interface for containers, images, builds, registries, networks, volumes, Swarm services, and plugins.

SetupDocsCommunityMaturity

Verdict Our `docker-trust` run installed 176 packages and built in 68 seconds, but its tests stopped with 3 of 7 packages failing to compile. That narrow failure does not undo Docker CLI's place as the standard client for Docker Engine, yet it blocks a clean recommendation for this commit to trust-plugin contributors.

Full review →
309

jira-cli

ankitpokhrel/jira-cli · ★ 6.0k16/7d · Go

JiraCLI puts common Jira work in a terminal, including finding issues, creating and cloning tickets, linking work, changing status, and exploring epics or sprints. It is meant to reduce trips through the Jira web interface, not replace every part of it.

SetupDocsCommunityMaturity

Verdict Our run built in 39 seconds and passed 26 of 26 tests. JiraCLI is an easy recommendation for terminal-heavy Jira users on Linux or macOS who want the daily issue loop to feel less cumbersome.

Full review →
310

likec4

likec4/likec4 · ★ 5.7k26/7d · TypeScript

LikeC4 is a text language and toolset for keeping software architecture diagrams beside the code they describe. It turns one model into interactive views, static sites, image exports, editor previews, generated components, and a queryable MCP server.

SetupDocsCommunityMaturity

Verdict Our LikeC4 run passed a 148-second build and 71-second test step after installing 1,608 packages, so the project earns trust but demands real workspace capacity. Use it when the architecture model itself should be reviewed, queried, and reused across several views.

Full review →
311

skills-manager

xingkongliang/skills-manager · ★ 4.8k180/7d · Rust

Skills Manager is a desktop app and CLI for keeping agent skills in one library, then deploying them to Claude Code, Codex, Cursor, Copilot, Gemini CLI, and dozens of other tools. Its main README is in English, with a full Chinese translation, and the interface supports both languages.

SetupDocsCommunityMaturity

Verdict Skills Manager is worth using once the same skill library has to serve several agents or computers. Its workspace model, shared Rust core, CLI, and conflict-aware Git backup address real maintenance pain, and our measured build passed.

Full review →
312

PI-Desktop

vastsa/PI-Desktop · ★ 3.8k2.3k/7d · TypeScript

PI-Desktop is a desktop workspace for running AI coding agents against local projects without tying the workflow to one editor or terminal. It keeps sessions, files, command output, diffs, previews, model settings, and permission decisions in one application while letting you bring your own model provider.

SetupDocsCommunityMaturity

Verdict Our PI-Desktop run installed 792 packages, built successfully, and passed all 22 tests, so the early preview is credible enough for a hands-on trial. It is a good fit for developers who want several agents, providers, and projects in one inspectable desktop.

Full review →
313

codex-desktop-linux

ilysenko/codex-desktop-linux · ★ 3.8k33/7d · JavaScript

ChatGPT Community for Linux is an unofficial distribution that verifies OpenAI's signed Linux ChatGPT package, wraps it for several Linux package systems, and adds optional community features. It gives Linux users Chat, Work, and Codex in a desktop app while keeping the unmodified application archive byte-identical when no optional patch is enabled.

SetupDocsCommunityMaturity

Verdict Our 3-CPU sandbox installed 253 packages and passed all 1,578 tests, which is unusually convincing evidence for an unofficial desktop repackager. Use it if you want ChatGPT's Linux desktop payload across several distributions and accept local builds, shared-profile rules, and upstream drift work.

Full review →
314

datadog-agent

DataDog/datadog-agent · ★ 3.7k4/7d · Go

Datadog Agent is the software that runs on hosts and clusters to collect metrics, traces, logs, processes, and security data for the Datadog service. This repository contains the source for Agent versions 6 and 7, including separate binaries for host, cluster, tracing, system, and security work.

SetupDocsCommunityMaturity

Verdict Our Datadog Agent run built in 7 seconds, but pytest ended after 8 seconds with 17 passes and 14 collection/setup errors, so the checkout did not earn a clean test result in a plain Python container. Use the released Agent if your organization already buys Datadog and wants its deepest host and cluster integration.

Full review →
315

lavish-axi

kunchenguid/lavish-axi · ★ 3.7k323/7d · JavaScript

Lavish AXI is a local browser editor for HTML files produced by coding agents. It lets a reviewer point at elements, select text, attach images, or edit Mermaid diagrams, then sends that feedback back to the agent through a command-line polling loop.

SetupDocsCommunityMaturity

Verdict Our Lavish AXI run installed 390 packages in 24 seconds and passed all 1,141 tests, making it an easy trial for developers who already review agent-made HTML. Use it for a local, visual feedback loop with Claude Code or Codex, especially when comments need to point at exact page elements.

Full review →
316

tuios

Gaurav-Gosain/tuios · ★ 3.7k37/7d · Go

TUIOS is a terminal multiplexer that makes command-line panes behave more like windows in a tiling desktop. It gives you nine workspaces, mouse and keyboard pane management, persistent sessions, layout automation, remote access, and status indicators for coding agents inside an ordinary terminal.

SetupDocsCommunityMaturity

Verdict Our TUIOS run built successfully and passed all 32 tests, but the test step took 248 seconds and does not cover every terminal protocol. It is a strong personal-workstation choice for developers who want a tiling desktop and visible coding-agent states inside a terminal.

Full review →
317

tgrep

microsoft/tgrep · ★ 3.1k692/7d · Rust

tgrep is a Rust search tool that builds a trigram index so repeated regex searches do not have to scan an entire large repository each time. It runs a local TCP server, keeps the index updated as files change, and is aimed at codebases where ordinary grep-style rescanning has become a noticeable delay.

SetupDocsCommunityMaturity

Verdict Our build finished in 29 seconds, and all 1,618 tests passed in 71 seconds, unusually convincing results for a young v1.0.4 developer tool. Use tgrep when repeated search latency in a genuinely large repository is a daily cost and running a local index service is acceptable.

Full review →
318

polkadot-sdk

paritytech/polkadot-sdk · ★ 2.8k1/7d · Rust

Polkadot SDK is the Rust codebase used to build Polkadot-compatible blockchains, parachains, runtimes, nodes, and cross-chain features. It combines the former Substrate, Polkadot, and Cumulus repositories, while the live Polkadot runtime itself is maintained separately by the Fellowship.

SetupDocsCommunityMaturity

Verdict Our 3-CPU sandbox spent 900 seconds on the build and another 900 seconds on tests without reaching either finish line, so Polkadot SDK is for teams that can fund serious Rust build infrastructure. Use it when FRAME, Cumulus, or XCM is part of the product requirement and a custom chain is worth the maintenance load.

Full review →
319

stremio-core

Stremio/stremio-core · ★ 2.4k · Rust

Stremio Core is the shared Rust engine behind Stremio's apps. It manages user and addon state, libraries, playback, notifications, and links so each platform can put its own interface on the same behavior. This is application infrastructure rather than a finished media player.

SetupDocsCommunityMaturity

Verdict Our Stremio Core run installed 131 packages and passed all 512 tests, so its Rust code is a credible base for teams building a Stremio-compatible client. Use it when Stremio's state model and addon protocol are requirements you already have.

Full review →
320

token-monitor

Javis603/token-monitor · ★ 2.2k146/7d · JavaScript

Token Monitor is a desktop dashboard for token use, estimated cost, account limits, and session activity across AI coding tools. It reads local records from tools such as Claude Code, Codex, Cursor, and OpenCode, then can combine summaries from several computers through a self-hosted hub.

SetupDocsCommunityMaturity

Verdict Our Token Monitor run installed 340 packages, passed all 3,674 tests, and still found 1 high-severity advisory, so it earns a trial with an audit review before wider deployment. It is the most useful fit for one developer juggling several coding subscriptions across a few machines.

Full review →
321

try-omarchy

themartiano/try-omarchy · ★ 2.1k310/7d · Swift

Try Omarchy packages the Omarchy Linux desktop as a Mac app for Apple Silicon. It gives Mac owners a persistent, hardware-accelerated Linux workspace without asking them to partition a disk or assemble the virtual machine themselves.

SetupDocsCommunityMaturity

Verdict Our run installed 35 packages and built in 35 seconds combined, but 1 of 30 tests failed, so Try Omarchy is easy to trial and still needs cautious evaluation. Use the signed app if you own an Apple Silicon Mac and specifically want Omarchy with Mac clipboard, camera, and folder integration.

Full review →
322

backpass

kunchenguid/backpass · ★ 1.0k113/7d · JavaScript

Backpass reads coding-agent transcripts from your machine and proposes small, evidence-backed changes to AGENTS.md, CLAUDE.md, and project skills. It is meant for teams that keep correcting the same agent mistakes by hand and want those corrections tied to real sessions before they reach a shared instruction file.

SetupDocsCommunityMaturity

Verdict Our Backpass run installed 77 packages in 15 seconds and all 472 tests passed, so the code is easy to trial on a supported machine. Use it if repeated agent mistakes already show up across several local sessions and every memory edit will receive human review.

Full review →
323

zoetrope

furkankly/zoetrope · ★ 87669/7d · Rust

Zoetrope turns Claude Code's local JSONL session transcripts into a live or replayable graph of the main agent, subagents, workflows, and tool calls. It runs as a terminal application or in a browser, helping developers inspect how an agent reached an outcome without uploading the transcript.

SetupDocsCommunityMaturity

Verdict Our build took 30 seconds and all 380 tests passed in 13 seconds, making Zoetrope unusually easy to trust for a first release. Install it if Claude Code subagents have made raw JSONL too hard to follow and local-only inspection matters.

Full review →
324

ai-usagebar

akitaonrails/ai-usagebar · ★ 51055/7d · Rust

AI Usage Bar puts the remaining limits and reset times for Claude, Codex, Copilot, OpenRouter, and other AI coding services in a desktop bar or terminal screen. It saves people with several subscriptions from opening each provider's billing page to decide which account can take the next job.

SetupDocsCommunityMaturity

Verdict Our AI Usage Bar run installed 255 packages, built successfully, and passed all 3,124 tests, which makes it an easy recommendation for developers juggling several AI plans on one machine. The catch is outside the Rust code: many providers expose usage through undocumented or private routes, so expect an occasional broken integration when a vendor changes its response.

Full review →
325

tech-interview-handbook

yangshun/tech-interview-handbook · ★ 143k171/7d · TypeScript

Tech Interview Handbook is a free English guide for software engineers preparing for hiring loops. It condenses coding practice, algorithms, resumes, behavioral interviews, and offer negotiation into a website and open repository for people who do not want to grind through hundreds of random questions.

SetupDocsCommunityMaturity

Verdict Our install added 1,342 packages and 850 MB in 36 seconds, but the build failed 6 seconds later on a Node.js export error. Read Tech Interview Handbook if you want a strong free map of the interview process, especially coding, resumes, and behavioral rounds.

Full review →
326

zed

zed-industries/zed · ★ 90k315/7d · Rust

Zed is a graphical code editor written largely in Rust, built around fast native rendering and live collaboration. It aims to cover everyday coding, language tooling, Git work, and AI agents without using a browser-based desktop shell.

SetupDocsCommunityMaturity

Verdict Our source build failed after 844 seconds, and the test attempt failed after 271 seconds at the same timed-out WebRTC download, so contributors need a more reliable dependency path than our fresh sandbox had. Zed is still a credible daily editor through its published builds, with current releases and unusually active development.

Full review →
327

system-design-101

ByteByteGoHq/system-design-101 · ★ 89k248/7d

System Design 101 is a visual, plain-English library of guides about APIs, networking, databases, architecture, and real-world systems. It helps developers build a broad mental map of system design topics and prepare for interviews without starting with a dense textbook.

SetupDocsCommunityMaturity

Verdict Our install finished in 22 seconds, but there was nothing to build or test. Use System Design 101 as a polished visual index and interview refresher, not as a runnable system or an authoritative specification.

Full review →
328

warp

warpdotdev/warp · ★ 65k137/7d · Rust

Warp is a Rust desktop and terminal interface that combines shell work, code navigation, a built-in coding agent, and support for external CLI agents such as Claude Code and Codex. This repository now contains Warp's open-source client codebase, while cloud-backed agent and sync features still connect to Warp services.

SetupDocsCommunityMaturity

Verdict Our Warp build took 860 seconds, and the test command failed with exit code 101 after 674 seconds without naming the failed case in the supplied tail. Download the packaged app if its terminal-plus-agent workflow appeals to you; building this Rust workspace is contributor work, not the sensible trial path.

Full review →
329

BMAD-METHOD

bmad-code-org/BMAD-METHOD · ★ 53k266/7d · JavaScript

BMAD-METHOD installs a structured software-delivery process into AI coding tools, including Claude Code. It turns an idea, bug, or specification into planning artifacts, implementation steps, and review passes so decisions survive beyond one chat.

SetupDocsCommunityMaturity

Verdict Our BMAD-METHOD run installed 786 packages in 26 seconds but failed its tests after 21 seconds, while npm audit reported 8 high-severity vulnerabilities. Try it if your recurring problem is lost context across planning, coding, and review, and you can own the method as part of the repository.

Full review →
330

tldraw

tldraw/tldraw · ★ 50k173/7d · TypeScript

tldraw is a source-available React SDK for putting an infinite drawing canvas inside a web application. It supplies the editor, default whiteboard interface, data model, multiplayer sync, and extension points needed to build diagramming, annotation, workflow, or visual AI products.

SetupDocsCommunityMaturity

Verdict Our tldraw install pulled 3,204 packages and occupied 3,364 MB, then the full build hit 900 seconds and the test run ended with 4 failed files. Use the SDK when its finished editor behavior can save a React team months of canvas work and the production license fits the budget.

Full review →
331

shannon

KeygraphHQ/shannon · ★ 48k157/7d · TypeScript

Shannon is an AI pentester that reads a web application's source, explores the running target, and attempts real exploits against its pages and APIs. It reports findings only when it can produce a proof of concept, with Markdown, JSON, and optional SARIF output.

SetupDocsCommunityMaturity

Verdict Our Shannon run installed 413 packages and built in 12 seconds, but the repository supplied no test target, so build success is the strongest automated result we can report. Use it as a second set of hands for an authorized staging pentest when reproducible exploitation matters.

Full review →
332

nvm

nvm-windows/nvm · ★ 48k81/7d · Go

NVM for Windows installs multiple Node.js versions on one Windows machine and switches the active version through a system symlink. It solves a common Windows development problem without depending on Node itself, since the command-line program is written in Go.

SetupDocsCommunityMaturity

Verdict Our Linux build completed in 48 seconds, but all 13 Go test targets failed during setup or compilation, so the source builds while its test path is not portable to our sandbox. On Windows, NVM for Windows remains a practical choice when you can give it administrator rights and a clean Node installation path.

Full review →
333

nvm

coreybutler/nvm-windows · ★ 48k70/7d · Go

NVM for Windows is a Go command-line utility for installing several Node.js versions on Windows and switching the active one through a system symlink. GitHub now redirects the original `coreybutler/nvm-windows` repository to `nvm-windows/nvm`; it is a separate project from the Unix-focused `nvm-sh/nvm`.

SetupDocsCommunityMaturity

Verdict Our Linux run built NVM for Windows in 1 second, then all 6 test packages failed because Windows syscall symbols were unavailable, so that failure describes platform coupling rather than a broken Windows release. Use it when you need a familiar installer and system-wide Node switching on Windows.

Full review →
334

iina

iina/iina · ★ 46k87/7d · Swift

IINA is a native macOS media player built around mpv, giving Mac users a polished way to play local video, audio, subtitles, playlists, chapters, and online media. It solves the gap between mpv's capable playback engine and the friendly, Mac-specific interface most people want for everyday viewing.

SetupDocsCommunityMaturity

Verdict Our 3-CPU, 8 GB Debian run produced no build or test result because IINA targets macOS and supplies no Dockerfile. Use IINA if you want mpv's playback depth in a thoughtful Mac interface, especially for subtitles, filters, custom controls, or picture-in-picture.

Full review →
335

herdr

herdrdev/herdr · ★ 39k2.0k/7d · Rust

Herdr is a persistent terminal workspace built around coding agents. It keeps agents in real panes, marks them working, blocked, done, idle, or unknown, and gives people and other agents a CLI and socket API for creating panes, sending prompts, and waiting for state changes.

SetupDocsCommunityMaturity

Verdict Our Herdr install fetched 174 packages in 32 seconds, but both build and test failed because `build.rs` could not execute Zig for `libghostty-vt`. Try the prebuilt v0.8.2 binary if agent-aware panes solve a real supervision problem, then test detection, detach, restart, and remote behavior with your own agents.

Full review →
336

Hyprland

hyprwm/Hyprland · ★ 39k99/7d · C++

Hyprland is a dynamic tiling Wayland compositor for Linux desktops. It manages windows, workspaces, displays, input, and visual effects for people who want far more control than a conventional desktop environment provides.

SetupDocsCommunityMaturity

Verdict Our measured subproject build finished in 7 seconds, but pytest exited 5 with 0 tests collected, so this checkout did not earn a clean test result. Hyprland is a compelling choice for experienced Linux users who actively want to shape their desktop and accept the maintenance that follows.

Full review →
337

serenity

SerenityOS/serenity · ★ 34k34/7d · C++

SerenityOS is a graphical Unix-like operating system built in one repository, including its kernel, desktop, browser, applications, libraries, and developer tools. It is meant for people who want to study and build an operating system with a late-1990s desktop style, rather than install a conventional Linux distribution.

SetupDocsCommunityMaturity

Verdict Our SerenityOS run installed 36 packages in 20 seconds and built the `Tests/ClangPlugins` Python project in 7 seconds; those results do not cover the operating system build. SerenityOS is an excellent codebase to study or join if compiling the toolchain and living on a moving ABI sound enjoyable.

Full review →
338

codex-plugin-cc

openai/codex-plugin-cc · ★ 33k273/7d · JavaScript

Codex Plugin for Claude Code lets you run Codex reviews and delegated coding tasks from a Claude Code session. It keeps the handoff, job status, and results inside Claude Code while using the Codex CLI already installed on your machine.

SetupDocsCommunityMaturity

Verdict Our sandbox installed 3 packages in 16 seconds and passed all 91 tests, but the build stopped after 9 seconds because the `codex` command was unavailable. Use this plugin if Claude Code is already your daily workspace and a second-agent review or resumable Codex handoff is worth the extra local machinery.

Full review →
339

oh-my-codex

Yeachan-Heo/oh-my-codex · ★ 33k130/7d · TypeScript

Oh My Codex, usually called OMX, is a workflow and runtime layer around OpenAI Codex CLI. It adds reusable planning and execution skills, persistent task state, worktree isolation, hooks, a terminal HUD, and coordinated agent teams for developers who find the stock CLI too bare.

SetupDocsCommunityMaturity

Verdict Our oh-my-codex test command hit the 900-second cap with 35 tests passed and none failed, while npm audit reported 6 known vulnerabilities. OMX is worth trying if Codex CLI is already central to your work and you want an opinionated, resumable operating system around it.

Full review →
340

sharp

lovell/sharp · ★ 33k · JavaScript

sharp is an image-processing library for JavaScript runtimes that turns large source images into smaller, web-friendly files. It handles resizing, format conversion, rotation, extraction, compositing, colour profiles and transparency, giving application developers a programmable alternative to shelling out to a separate image utility.

SetupDocsCommunityMaturity

Verdict Our source build failed in 4 seconds, despite dependency installation succeeding in 30 seconds with 0 known audit vulnerabilities. Use sharp for Node.js image pipelines when its focused API and libvips foundation match the job, but verify your exact runtime and deployment path early.

Full review →
341

egui

emilk/egui · ★ 31k88/7d · Rust

egui is a Rust library for building interactive graphical interfaces from the same code on native platforms, the web, and supported game engines. Its immediate-mode approach lets an application describe the current interface each frame, which is useful when a conventional native-looking toolkit would add more structure than the job needs.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 640 seconds, but 2 of 152 tests failed. egui is an easy recommendation for Rust-native tools, engine overlays, and custom interfaces where portability matters more than an operating system's native look.

Full review →
342

notepad-plus-plus

notepad-plus-plus/notepad-plus-plus · ★ 29k55/7d · C++

Notepad++ is a free Windows text and source-code editor whose codebase is primarily C++; the supplied README is in English. It replaces the basic Notepad experience with support for programming and natural languages, solving the need for a focused editor without requiring a full development environment.

SetupDocsCommunityMaturity

Verdict We could not run Notepad++ in our Debian sandbox: our harness had no supported C++ ecosystem and the repository supplied no Dockerfile. That result limits what we can certify, but it also matches the project's explicit focus on Windows rather than containerized deployment.

Full review →
343

CloudflareSpeedTest

XIU2/CloudflareSpeedTest · ★ 29k58/7d · Go

CloudflareSpeedTest is a Chinese-language command-line tool that samples CDN addresses, measures latency and download speed, and writes the best results to a CSV file. The project has no first-party English README, though its README links to an unofficial English-language fork.

SetupDocsCommunityMaturity

Verdict Our CloudflareSpeedTest run installed 9 packages, built in 26 seconds, and completed `go test` in 4 seconds, but the test command exercised 0 tests. It is a focused, easy-to-build utility for Chinese-speaking operators who understand that results describe one connection at one moment.

Full review →
344

system-design-academy

systemdesign42/system-design-academy · ★ 29k91/7d

System Design Academy is a curated index of system-design case studies, fundamentals, interview material, AI engineering articles, and software white papers. It solves the discovery problem for engineers who want one browsable starting point instead of hunting across many publications, but it is a reading list, not an application or executable toolkit.

SetupDocsCommunityMaturity

Verdict Our run executed 0 install, build, or test commands because the repository had no detected language, supported ecosystem, or Dockerfile. Use System Design Academy as a high-volume discovery page, especially if company architecture stories help you choose what to read next.

Full review →
345

open-code-review

alibaba/open-code-review · ★ 29k6.4k/7d · Go

OpenCodeReview is an AI code-review CLI that selects changed files, groups related code, applies path-specific rules, gives an LLM tools to inspect the repository, and returns line-level findings. It can review working-tree changes, commit ranges, or whole files, and it integrates with CI, MCP, Claude Code, Codex, Cursor, and other coding agents.

SetupDocsCommunityMaturity

Verdict Our OpenCodeReview checkout could not install because its `0.0.0` package version produced a GitHub asset URL that returned 404 after 12 seconds. Try the v1.10.1 release binary or published npm package instead, then compare its findings with human review on your own changes.

Full review →
346

shields

badges/shields · ★ 27k · JavaScript

Shields.io turns project facts such as build status, package versions, coverage, downloads, and custom labels into small SVG or raster badges for READMEs and web pages. It solves the otherwise repetitive job of fetching those facts and rendering them in a consistent, readable format, while also providing a badge-generation library and a published design specification.

SetupDocsCommunityMaturity

Verdict Our build completed in 180 seconds, but 31 of 202 Mocha tests failed. Use the hosted service confidently for ordinary README badges, because its breadth, familiar output, and same-day repository activity make it the practical default.

Full review →
347

tree-sitter

tree-sitter/tree-sitter · ★ 27k63/7d · Rust

Tree-sitter is a parser generator and incremental parsing library for tools that need a syntax tree while code is being edited. It lets editors, analyzers, search tools, and refactoring systems update only the affected parts of a parse and still return a useful tree around syntax errors.

SetupDocsCommunityMaturity

Verdict Our Tree-sitter build finished in 60 seconds, but 266 of 305 tests failed in 9 seconds with no cause visible in the supplied log tail. The architecture remains a strong fit for editors and structural code tools, and the current release and binding ecosystem show mature adoption.

Full review →
348

GDevelop

4ian/GDevelop · ★ 27k210/7d · JavaScript

GDevelop is an open-source game editor and runtime for making 2D, 3D, and multiplayer games with an event system instead of writing the whole game in code. It publishes to the web, desktop, and mobile, while still giving contributors access to the JavaScript engine, React and Electron editor, extensions, and C++ core.

SetupDocsCommunityMaturity

Verdict Our GDevelop engine run installed 202 packages in 176 seconds and built in 6 seconds, but its test command stopped because `karma` was not found. The finished editor remains a sensible choice for visual game makers who want ownership of exported games and room to add code.

Full review →
349

NativeScript

NativeScript/NativeScript · ★ 26k6/7d · TypeScript

NativeScript is a TypeScript and JavaScript toolkit for building native iOS, Android, and visionOS applications while calling platform APIs directly. It solves the problem of sharing application logic and familiar web-language skills without limiting developers to browser-based interfaces.

SetupDocsCommunityMaturity

Verdict Our clean-container run installed 1,515 packages in 91 seconds and reported 0 known vulnerabilities. NativeScript is a serious choice when direct native API access and JavaScript or TypeScript are both requirements, especially if you value its range of framework integrations.

Full review →
350

coc.nvim

neoclide/coc.nvim · ★ 25k · TypeScript

coc.nvim is an IntelliSense engine that gives Vim and Neovim editor features similar to those in VS Code. It handles completion, diagnostics, code navigation, refactoring, and extensions through a Node.js service, helping terminal-editor users get richer coding assistance without changing editors.

SetupDocsCommunityMaturity

Verdict Our install finished in 15 seconds and the build in 6 seconds, but the test run still exited 1. coc.nvim remains a serious choice for Vim or Neovim users who want a broad, integrated editing layer and accept Node.js as part of the stack.

Full review →
351

spacemacs

syl20bnr/spacemacs · ★ 25k1/7d · Emacs Lisp

Spacemacs is a ready-made Emacs configuration that organizes editing commands behind memorable key sequences and lets people use Vim or Emacs input styles. It solves the blank-canvas problem of building an Emacs setup by supplying a UI, conventions, documentation, and configuration layers for common programming work.

SetupDocsCommunityMaturity

Verdict Our sandbox produced no install, build, or test result for commit bb9691a because Emacs Lisp is outside the runner and the repository has no Dockerfile. Spacemacs is still worth trying if you want Emacs and Vim habits in one maintained configuration, but its rolling updates and first-launch bootstrap demand more trust than a pinned editor package.

Full review →
352

beekeeper-studio

beekeeper-studio/beekeeper-studio · ★ 24k · TypeScript

Beekeeper Studio is a desktop SQL editor and database manager for Windows, macOS, and Linux. Its free Community Edition covers 11 database families, while another group of drivers and features requires a paid edition.

SetupDocsCommunityMaturity

Verdict Our Beekeeper Studio checkout installed 3,626 packages and occupied 1,738 MB, but it exposed no build or test target, so contributors get an installed tree without a project-level verification step. The downloadable app is worth trying if its 11 free database families cover your work and you prefer a restrained desktop interface.

Full review →
353

superfile

yorukot/superfile · ★ 23k120/7d · Go

superfile is a terminal file manager that puts browsing, previews, metadata, file operations, and a command bar into a panel-based interface. It is for people who want more visual context than shell commands provide while keeping file work inside the terminal.

SetupDocsCommunityMaturity

Verdict Our superfile build succeeded in 40 seconds, but `go test` reported 22 passes and 7 failures out of 29, all listed under the zoxide UI package. The release binaries are worth trying for a local Linux or macOS workflow because the panels and file operations solve a real terminal usability problem.

Full review →
354

micropython

micropython/micropython · ★ 22k16/7d · C

MicroPython is a compact implementation of Python 3.x for microcontrollers, embedded systems, and other constrained devices. It lets developers write familiar Python while retaining direct access to hardware such as GPIO, timers, buses, Bluetooth, and USB.

SetupDocsCommunityMaturity

Verdict Our build failed after 9 seconds, and our 19-second test run passed 0 tests because all 44 collected paths ended in setup or collection errors. That result makes MicroPython a poor choice if you expect a generic Python repository workflow, but it does not erase the project's clear embedded focus or active maintenance.

Full review →
355

theia

eclipse-theia/theia · ★ 22k15/7d · TypeScript

Eclipse Theia is a framework for building your own browser-based or desktop coding environment, rather than a finished editor you simply download and use. It gives product teams an extensible IDE foundation, including support for the VS Code extension protocol, while keeping the project under vendor-neutral open-source governance.

SetupDocsCommunityMaturity

Verdict Our npm install failed after 275 seconds at the native-keymap 2.5.0 node-gyp rebuild, so Theia is not a low-friction trial in the measured environment. Use it when an IDE is part of your product and architectural control justifies owning a large framework.

Full review →
356

Ciphey

bee-san/Ciphey · ★ 22k17/7d · Rust

Ciphey is a Rust command-line tool and library that tries to turn unknown encoded or classically encrypted text into readable output. It searches through decoders, checks candidate text, and reports the path it used, saving the trial-and-error work common in capture-the-flag puzzles and data triage.

SetupDocsCommunityMaturity

Verdict Our Ciphey run installed 335 packages, built in 89 seconds, and passed all 612 tests in 150 seconds, making v0.12.1 a credible CTF and decoding assistant. Use it to generate and rank plausible decoding paths, then verify the result yourself.

Full review →
357

gofr

gofr-dev/gofr · ★ 21k · Go

GoFr is an opinionated Go framework for building microservices, with built-in HTTP, gRPC, configuration, logging, tracing, metrics, and pub/sub. It solves the problem of assembling and wiring the same Kubernetes-friendly service boilerplate in every new Go service.

SetupDocsCommunityMaturity

Verdict Our clean checkout of gofr-dev/gofr compiled successfully, but its own test suite failed: 95 passed and 12 failed out of 107 total. GoFr is a real option if you want a batteries-included Go microservice framework and are willing to accept some test-level rough edges..

Full review →
358

triton

triton-lang/triton · ★ 20k61/7d · MLIR

Triton is a Python-based language and compiler for writing custom deep-learning kernels without dropping all the way into CUDA. It lets specialists describe tiled tensor work at a higher level, then compiles that work for supported GPUs.

SetupDocsCommunityMaturity

Verdict Our Triton source install reached the 900-second cap and ended with a failed CMake build, so this is specialist compiler work even though the user wheel installs with pip. Choose Triton when a measured model bottleneck justifies a custom Linux GPU kernel and your team can test it on each target.

Full review →
359

migrate

golang-migrate/migrate · ★ 19k14/7d · Go

golang-migrate/migrate applies ordered database changes from files, embedded assets, or remote storage. You can run it as a command-line program in any stack or import its version 4 Go library when migration control belongs inside an application.

SetupDocsCommunityMaturity

Verdict Our golang-migrate run installed 457 packages in 68 seconds and built in 174 seconds, but 18 of 68 Go package results failed. The released CLI remains a sensible choice for teams that already write and review explicit migrations and need broad driver coverage.

Full review →
360

webvm

leaningtech/webvm · ★ 17k9/7d · JavaScript

WebVM runs an x86 Linux environment inside a web browser, using WebAssembly instead of a server-side virtual machine. It gives people a disposable Debian terminal or graphical Linux workspace on devices where installing a normal VM is inconvenient or impossible.

SetupDocsCommunityMaturity

Verdict Our WebVM checkout installed 233 packages and built in 61 seconds combined, but it offered no test target for checking the virtualization layer. Use it for a bounded browser lab, demo, or classroom where you can test every required command yourself.

Full review →
361

gpui-kit

longbridge/gpui-component · ★ 14k225/7d · Rust

The repository formerly called GPUI Component now resolves to GPUI Kit, a Rust framework for native desktop interfaces. It combines styled controls, lower-level UI behavior, data-heavy widgets, and optional JavaScript extensions so teams do not have to assemble those parts from separate libraries.

SetupDocsCommunityMaturity

Verdict Our GPUI Kit build ran for 431 seconds before `xkbcommon-x11` stopped the linker, and its tests hit the same missing library after 14 seconds. Use it for a Rust desktop product whose hard parts are tables, text editing, docking, and shared input behavior, provided your team can own native setup and accessibility verification.

Full review →
362

nuclei-templates

projectdiscovery/nuclei-templates · ★ 13k38/7d · JavaScript

Nuclei Templates is the shared rule library used by ProjectDiscovery's Nuclei security scanner. It gives defenders and authorized testers ready-made checks for exposed services, known vulnerabilities, weak configurations, and other findings, while the separate Nuclei project supplies the scanning engine.

SetupDocsCommunityMaturity

Verdict Our sandbox produced no install, build, or test result for commit ed8fed8 because this is a scanner-content repository without a supported runnable ecosystem or Dockerfile. Use Nuclei Templates when your team already understands Nuclei and treats matches as leads to verify, because the 11,997-file catalog is current and useful but individual checks can still be wrong.

Full review →
363

iroh

n0-computer/iroh · ★ 13k52/7d · Rust

Iroh is a Rust networking library that lets an application connect to a peer by public key instead of managing changing IP addresses. It tries a direct QUIC path, handles NAT traversal, and can fall back to relays when a direct connection does not work.

SetupDocsCommunityMaturity

Verdict Our Iroh checkout installed 461 packages, but both the 20-second build and 9-second test run failed when Clang rejected the configured `lld` linker selection. The public-key dialing model is attractive for Rust applications that truly need to cross NATs, and the separate relay and protocol crates give serious teams room to own more of the stack.

Full review →
364

garden-skills

ConardLi/garden-skills · ★ 12k172/7d · CSS

Garden Skills is an English-first collection of five instruction packs for Claude Code, Codex, Cursor, and other coding agents, with Chinese and Japanese versions of the main README. It covers web design, browser-recorded presentations, GPT Image 2 prompting, knowledge retrieval, and turning source material into styled articles.

SetupDocsCommunityMaturity

Verdict Our Garden Skills run installed 0 packages in 21 seconds, but it had no build or test target, so adoption depends on reading and trialing the individual instructions rather than trusting a passing suite. Use it when one of the five workflows matches a repeated job and your team is comfortable reviewing agent output.

Full review →
365

sliver

BishopFox/sliver · ★ 12k37/7d · Go

Sliver is an open-source command-and-control framework for authorized red-team and adversary-emulation work. It helps security teams create and operate test implants across Windows, macOS, and Linux so they can assess whether defenses detect realistic attacker behavior.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 13 seconds, but 66 of 137 tested packages failed in the same sandbox. Sliver is a serious, unusually capable choice for experienced red teams that want cross-platform implants and several C2 transports in one open-source system.

Full review →
366

fff

dmtrKovalenko/fff · ★ 11k81/7d · Rust

FFF is a persistent file and content search engine for editors, coding agents, and long-running developer tools. It combines typo-tolerant matching, an in-memory index, filesystem watching, Git status, and frecency ranking, then exposes them through MCP, Neovim, Pi, Rust, Python, Node, Bun, and C interfaces.

SetupDocsCommunityMaturity

Verdict Our npm check completed in 6 seconds with 0 installed packages and 0 known vulnerabilities, but it built and tested none of FFF's Rust or native deliverables. FFF is a strong candidate for an editor or coding agent that searches one repository many times and can validate its platform-specific binary.

Full review →
367

wgcf

ViRb3/wgcf · ★ 8.7k17/7d · Go

wgcf is an unofficial command-line client that registers a Cloudflare WARP account and turns it into a standard WireGuard configuration. It lets servers, routers, and other headless systems use WARP without Cloudflare's official app.

SetupDocsCommunityMaturity

Verdict Our wgcf run installed 63 packages in 35 seconds, built in 39 seconds, and passed all 3 tests in 11 seconds. That clean result makes it an easy trial for a headless system that specifically needs a WARP-backed WireGuard profile.

Full review →
368

ui-skills

ibelick/ui-skills · ★ 8.6k304/7d · TypeScript

UI Skills is a public registry of instruction files for agents doing interface design and frontend work. Its website, CLI, and Model Context Protocol endpoint help an agent find guidance on accessibility, motion, visual cleanup, design systems, and framework-specific UI tasks.

SetupDocsCommunityMaturity

Verdict Our ui-skills install used 687 MB, then built in 42 seconds and passed all 58 tests in 13 seconds with 0 known vulnerabilities. Use the registry when an agent needs a sharp checklist for one UI concern, especially accessibility or motion performance.

Full review →
369

plugins

cursor/plugins · ★ 7.9k743/7d · TypeScript

Cursor Plugins is the official public marketplace repository and format reference for extending the Cursor coding agent. Its catalog packages skills, rules, commands, hooks, subagents, and MCP connections for code review, memory, orchestration, browser work, email, CRM, and other developer workflows.

SetupDocsCommunityMaturity

Verdict Cursor Plugins is the right first stop for extending Cursor because it combines an official catalog with source you can inspect before granting access. The developer-focused plugins are useful and the manifest examples are valuable, but marketplace trust is not the same as uniform runtime safety.

Full review →
370

dsh-web

zhu1090093659/dsh-web-ui · ★ 7.6k372/7d · TypeScript

The project is primarily documented in Chinese, and a maintained English README is available as `README.en.md`. Now named dsh-web, it is a collection of plugins for the DeepSeek Harness web and desktop interfaces, adding a task board, remote access, SSH tools, skins, Git views, image understanding, and recovery utilities without modifying the DSH source.

SetupDocsCommunityMaturity

Verdict Our dsh-web-ui run built in 32 seconds, but its recursive test command stopped after `dsh-ssh` reported 139 of 140 tests passing. Install individual plugins first if one feature solves the problem; the full bundle makes sense for committed DSH users who will test profile migration and plugin interactions.

Full review →
371

whatsmeow

tulir/whatsmeow · ★ 7.3k55/7d · Go

whatsmeow is a Go library that connects an application to WhatsApp's web multi-device system. It handles linked-device login, messages, media, group operations, receipts, and events, leaving you to build the service around those pieces.

SetupDocsCommunityMaturity

Verdict Our whatsmeow run installed 39 packages, built successfully, and passed 2 of 2 tests in 71 seconds total, which makes the code easy to trial but leaves live WhatsApp behavior untested. It is the strongest fit for Go teams that need direct linked-device access and can own storage, upgrades, and account canaries.

Full review →
372

skills

jakubkrehel/skills · ★ 6.7k664/7d · Markdown

Jakub Krehel's Skills is a set of 11 English-language agent instructions for reviewing and improving web interfaces. It covers UI details, typography, color, accessibility, layout, product copy, change reviews, interface explanations, component stress tests, and design variants.

SetupDocsCommunityMaturity

Verdict Our sandbox could not execute commit 267330e because it is a Markdown skill collection with no supported ecosystem or Dockerfile, so adoption rests on reviewing the instructions and testing them in your own agent. The 11 skills are unusually specific about interface judgment and worth trying one at a time.

Full review →
373

actors

rivet-dev/rivet · ★ 6.1k11/7d · Rust

Rivet is an actor framework and orchestration engine for stateful services such as AI agents, collaborative documents, chat rooms, scheduled work, and per-tenant databases. Each actor keeps state near its code, receives queued messages, streams events over WebSockets, and can sleep when idle while its data persists.

SetupDocsCommunityMaturity

Verdict Our Rivet checkout installed 2,928 packages and used 3,846 MB, then both the 15-second build and 20-second test run failed in example workspaces. The library is still worth a focused proof of concept for TypeScript teams that genuinely need actor state, queues, schedules, and live connections.

Full review →
374

agentsview

kenn-io/agentsview · ★ 5.9k76/7d · Go

AgentsView is a local session browser and usage tracker for Claude Code, Codex, Gemini CLI, Copilot, and dozens of other coding agents. It reads their on-disk histories into SQLite so you can search prompts and tool calls, inspect recent edits, and estimate token costs without sending the transcripts to a hosted dashboard.

SetupDocsCommunityMaturity

Verdict Our AgentsView run built in 3 seconds but only 17 of 46 Go test packages passed, so the current checkout needs investigation before team deployment. It is useful for one developer who wants local search across Claude Code, Codex, and other agents, especially with the server kept on loopback.

Full review →
375

anchor

otter-sec/anchor · ★ 5.1k4/7d · Rust

Anchor is a Rust framework and command-line workspace for writing Solana programs, generating interface descriptions, and producing Rust or TypeScript clients. It reduces repeated account-validation and serialization code, while keeping programs inside Solana's execution rules.

SetupDocsCommunityMaturity

Verdict Our Anchor checkout completed its npm install path in 9 seconds with 0 packages added, but exposed no build or test target for the harness. That result is too shallow to certify a Solana framework, so adoption should depend on a real Rust, local-validator, and adversarial-test run.

Full review →
376

Tech-OA-Interview-Questions

perixtar/Tech-OA-Interview-Questions · ★ 5.1k206/7d · Python

Tech OA & Interview Questions is a community-maintained index of software engineering assessment and interview prompts, organized by company and format. It helps candidates find recent coding, SQL, system design, object-oriented design, and AI coding practice without searching scattered candidate reports.

SetupDocsCommunityMaturity

Verdict Our build finished in 4 seconds after a 12-second install, but there was no test target to establish correctness. Use this as a timely discovery board, especially when you know the company you are targeting, not as your only interview curriculum or as a dependable software dependency.

Full review →
377

cli

entireio/cli · ★ 5.1k20/7d · Go

Entire CLI records coding-agent sessions and links them to the Git commits they produced. It gives a team a searchable explanation of why code changed and a checkpoint from which an agent session can be resumed.

SetupDocsCommunityMaturity

Verdict Our Entire CLI run built in 84 seconds, but 3 of 85 test packages failed, so the tested commit is easy to compile and not clean enough to adopt without reproducing those failures. Use it when agent-session provenance is worth adding hooks and a checkpoint branch to every repository.

Full review →
378

devspace

Waishnav/devspace · ★ 4.8k198/7d · TypeScript

DevSpace is a self-hosted MCP server that lets ChatGPT and coding agents read, edit, search, and run commands in selected local projects. It provides a Codex-style workspace through a public HTTPS tunnel you control, with OAuth approval and an owner password before a client can connect.

SetupDocsCommunityMaturity

Verdict Our DevSpace run installed 281 packages and used 975 MB, then passed all 6 tests, while npm audit still reported 14 advisories, including 6 high severity. Use it if ChatGPT-based local coding is worth operating a tunnel and granting a trusted client shell access under a deliberately limited user account.

Full review →
379

teamai-cli

Tencent/teamai-cli · ★ 4.6k1.9k/7d · TypeScript

TeamAI CLI is a Git-backed distribution and memory layer for teams that use several coding agents. It sends shared skills, rules, documents, hooks, MCP settings, and selected session learnings into Claude Code, Codex, Cursor, and other supported tools so each agent starts with the team's working conventions.

SetupDocsCommunityMaturity

Verdict Our TeamAI CLI run installed 360 packages and used 895 MB, then passed all 2,729 tests, so the implementation is ready for a serious pilot despite its v0.23.0 label. Use it when agent configuration already belongs in your team's Git review process and several clients need the same instructions.

Full review →
380

nono

nolabs-ai/nono · ★ 4.1k87/7d · Rust

nono runs coding agents and the tools they call inside a local operating-system sandbox, limiting which files, networks, commands, and credentials each process can reach. It gives teams a reviewable policy outside the agent's prompt, without requiring a container or virtual machine for each session.

SetupDocsCommunityMaturity

Verdict Our nono run built in 363 seconds and passed 6,112 of 6,114 tests, but the failed suite and the project's own warning against production use keep it out of high-stakes deployment today. It is still one of the more thoughtful ways to reduce a local coding agent's reach, especially when delegated tools and API credentials need different rules.

Full review →
381

dora

dora-rs/dora · ★ 3.9k · Rust

Dora is a Rust runtime for building robots and AI systems from small programs connected as a dataflow graph. It starts and monitors nodes written in Rust, Python, C, or C++, moves Arrow data between them on one or several machines, and provides one CLI for running, inspecting, recording, and replaying the system.

SetupDocsCommunityMaturity

Verdict Our Dora test command hit the 900-second cap with 2 tests passed and 0 failed while Cargo was still compiling, despite a 19-second install and 1-second build. Dora deserves a trial for a new mixed-language robot because its graph, Arrow messages, record/replay tools, and 1.x node-API promise fit together well.

Full review →
382

rhwp

edwardkim/rhwp · ★ 3.8k15/7d · Rust

rhwp is a Korean-first Rust and WebAssembly viewer, editor, converter, CLI, and MCP server for HWP, HWPX, and a limited set of HML documents. A full English README exists, but the project says its roadmap and working documentation use Korean as the source of truth.

SetupDocsCommunityMaturity

Verdict Our rhwp build completed in 168 seconds and its test run recorded 3,880 passes with zero failures before the 900-second cap, which is strong evidence for a v0.8.4 project but not a completed suite. It is the most ambitious open-source choice here for web editing, conversion, and agent access to Korean HWP files.

Full review →
383

go-modern-guidelines

JetBrains/go-modern-guidelines · ★ 3.5k199/7d · Go

Modern Go Guidelines is a version-aware instruction set and small CLI that tells coding agents which Go idioms fit a project's declared Go version. It packages the advice for Junie, Claude Code, Codex, Cursor, and other skill-aware agents, addressing the stale patterns that models often copy from older training examples.

SetupDocsCommunityMaturity

Verdict Our run installed 3 packages and passed all 6 tests in 32 seconds total, so trying Modern Go Guidelines is cheap for a Go team that already uses coding agents. Use it as a version-aware reviewer prompt, then check every diff with the compiler, tests, and human judgment.

Full review →
384

tuicr

agavra/tuicr · ★ 3.1k60/7d · Rust

tuicr is a terminal interface for reviewing code with Vim-style navigation, inline comments, and saved review progress. It reads Git, Jujutsu, or Mercurial changes and can send a finished review to GitHub, GitLab, Bitbucket, or Azure DevOps, or export it as Markdown for another tool.

SetupDocsCommunityMaturity

Verdict Our tuicr run built successfully, but 20 of 3,104 tests failed, so Mercurial users should reproduce those failures before trusting it for daily review. Git and hosted-forge users still get an unusually complete keyboard-driven review flow, including persisted progress and real inline submission.

Full review →
385

takumi

kane50613/takumi · ★ 3.0k51/7d · Rust

Takumi is a Rust rendering engine that turns JSX, HTML, CSS, or a JSON node tree into social images, animations, SVG, and paged PDF without launching Chrome. JavaScript packages run through native bindings or WebAssembly, while Rust applications can embed the engine as a crate.

SetupDocsCommunityMaturity

Verdict Our Takumi checkout installed 950 packages and used 1,687 MB, then its root build and example-heavy test command both failed within 12 seconds. The published packages still deserve a focused trial for browserless OG images or structured PDFs because the API and supported output range are unusually broad.

Full review →
386

SenPaiScanner

MatinSenPai/SenPaiScanner · ★ 2.4k5/7d · Go

SenPai Scanner finds reachable Cloudflare edge addresses, ranks them, and can test the best candidates through a VLESS, Trojan, or VMess proxy configuration. It replaces blind IP guessing with discovery, end-to-end validation, speed testing, and client exports across desktop, Android, and terminal interfaces.

SetupDocsCommunityMaturity

Verdict Our SenPai Scanner run installed 251 Go packages, built in 75 seconds, and passed all 7 tests in 25 seconds. It is a good first choice when the whole job is discovering and exporting usable Cloudflare endpoints on the network in front of you.

Full review →
387

printing-press-library

mvanhorn/printing-press-library · ★ 2.0k23/7d · Go

Printing Press Library is a catalog and installer for hundreds of focused command-line tools generated for software agents. It lets Claude Code, Codex, OpenClaw, Hermes, and other skill-aware agents discover a service-specific CLI, install its Go binary and companion skill, then call it with less documentation in the prompt.

SetupDocsCommunityMaturity

Verdict Our Printing Press Library test run failed in 8 seconds because its compiled test glob was not found, even though installation and build both succeeded and npm audit found 0 vulnerabilities. The catalog is worth using as a discovery layer for agent tools, especially when one focused CLI replaces repeated API explanation.

Full review →
388

treehouse

kunchenguid/treehouse · ★ 1.7k65/7d · Go

Treehouse keeps a reusable pool of Git worktrees so parallel coding sessions get separate directories without cloning the repository again. It hands out a clean worktree in a subshell or through a durable lease, then checks and resets that slot before putting it back into the pool.

SetupDocsCommunityMaturity

Verdict Our Treehouse build passed in 111 seconds and all 10 Go tests passed in 232 seconds. It is a good fit for developers repeatedly assigning one repository to several agents, especially when dependency caches are expensive and lease-aware automation is possible.

Full review →
389

BrewUI

Homebrew/BrewUI · ★ 1.3k · Swift

BrewUI is Homebrew's official macOS app for finding, installing, updating, and removing formulae and casks without typing commands. It still runs the `brew` CLI underneath and shows its work in a console, so the interface does not hide what changed on your Mac.

SetupDocsCommunityMaturity

Verdict BrewUI requires macOS Tahoe 26 or newer, and our Debian lab could not run its Swift code, so our recommendation rests on the documented workflow and current issue record rather than a hands-on result. Try it if you already use Homebrew and want a first-party GUI that keeps command output visible.

Full review →
390

sofka

nklmilojevic/sofka · ★ 1.2k226/7d · Rust

Sofka is a Rust terminal interface for watching and operating Kubernetes clusters without constantly assembling kubectl commands. It puts standard resources, custom resources, logs, events, incident clues, port-forwards, and common Flux CD or Argo CD actions into one keyboard-driven screen. Its central idea is one generic object pipeline, so an unfamiliar CRD remains usable even without a purpose-built view.

SetupDocsCommunityMaturity

Verdict Our build completed in 151 seconds, but 4 of 1,362 tests failed. Sofka is already unusually capable for a v0.24.4 Kubernetes TUI, especially around arbitrary CRDs, native GitOps actions, and enforceable read-only contexts.

Full review →
391

herdr-reviewr

persiyanov/herdr-reviewr · ★ 70851/7d · Rust

herdr-reviewr is a terminal code-review pane built for the Herdr agent workspace manager. It lets you inspect an agent's diff, attach comments to lines, and send that feedback back to the agent without switching to a browser or editor.

SetupDocsCommunityMaturity

Verdict Our herdr-reviewr run built in 181 seconds, but 46 of 1,330 tests failed, so source adopters should require a clean run on their own Herdr setup. The prebuilt plugin is still worth trying for a terminal-first Herdr user because its four diff scopes and line-comment loop solve a specific interruption.

Full review →
392

purple

erickochen/purple · ★ 69315/7d · Rust

Purple is a terminal SSH manager for macOS and Linux that edits the user's existing SSH config, syncs hosts from 18 infrastructure providers, and puts file transfer, tunnels, remote containers, saved commands, keys, and Vault certificates in one interface. It also exposes an MCP server so AI clients can inspect or operate the same hosts.

SetupDocsCommunityMaturity

Verdict Our Purple run built in 97 seconds and passed 5,834 of 5,840 tests, but all 6 failures hit Vault SSH proxy-chain behavior, exactly where a wrong hop can block or misdirect access. It is worth adopting for a terminal-heavy operator who wants local SSH inventory plus remote container and tunnel tools, provided the config is backed up and Vault proxy paths are retested.

Full review →
393

omnyssh

timhartmann7/omnyssh · ★ 687 · Rust

OmnySSH is an open-source SSH manager with both a desktop app and a terminal interface. It puts host status, terminal tabs, SFTP transfers, reusable command snippets, SSH config imports, and key setup in one local application without requiring an account.

SetupDocsCommunityMaturity

Verdict Our OmnySSH build passed in 114 seconds, and all 886 tests passed in 42 seconds after a 14-second install. That makes the small Rust workspace easy to trust as code, while open password-storage and release-verification reports still need a security-conscious buyer's attention.

Full review →
394

tabby

Eugeny/tabby · ★ 75k · TypeScript

Tabby is a configurable terminal emulator that also puts SSH, Telnet, SFTP, and serial connections in one desktop application. It solves the hassle of juggling a local terminal, a separate remote-connection manager, and extra utilities across Windows, macOS, and Linux.

SetupDocsCommunityMaturity

Verdict Our source install failed after 92 seconds, so Tabby is not an automatic recommendation for teams that insist on reproducible builds from a fresh checkout. As a packaged desktop terminal, however, its combination of local shells, SSH, SFTP, Telnet, and serial support is unusually practical.

Full review →
395

act

nektos/act · ★ 72k · Go

Act runs GitHub Actions workflow files on your own machine, so you can check a job before pushing a commit. It maps workflow jobs to local Docker containers, supplies GitHub-like variables, and accepts event data and secrets. It is useful feedback, but it does not reproduce every feature or hosted-runner detail.

SetupDocsCommunityMaturity

Verdict Our Act build passed in 48 seconds, but only 4 of 7 test results passed and the 106-second suite exited 1. Use it for quick checks on container-friendly workflows, especially while editing YAML or an action.

Full review →
396

v2rayNG

2dust/v2rayNG · ★ 63k332/7d · Kotlin

v2rayNG is a Kotlin Android client that connects through either the Xray or v2fly core, giving users a mobile interface for V2Ray-compatible proxy setups. Its README provides both English and Chinese instructions, including downloads, routing-data management, signature checks, and developer setup. It solves the client-side configuration problem, but it does not provide a proxy server or subscription by itself.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 62 seconds, but our test command found 0 tests. v2rayNG is a strong choice for experienced Android users who already understand V2Ray-style profiles and value access to both Xray and v2fly cores.

Full review →
397

bevy

bevyengine/bevy · ★ 48k94/7d · Rust

Bevy is a Rust game engine for building 2D and 3D games around an Entity Component System. It gives programmers rendering, input, audio, scenes, UI, and an application framework without requiring a commercial editor or proprietary runtime.

SetupDocsCommunityMaturity

Verdict Our Bevy run installed 521 packages, built in 473 seconds, and was still compiling tests at the 900-second cap, so adopting it means accepting a large Rust workspace and slow clean checks. Use Bevy when an all-Rust, code-first engine is worth frequent API migrations and you can build your own production tooling.

Full review →
398

nw.js

nwjs/nw.js · ★ 41k · JavaScript

NW.js is a desktop application runtime that puts Chromium and Node.js in the same process, letting HTML and JavaScript interfaces call Node modules directly. It gives web developers a route to Windows, macOS, and Linux apps without replacing the interface with a native toolkit.

SetupDocsCommunityMaturity

Verdict Our NW.js checkout installed 50 packages in 20 seconds, then its build exited 1 and pytest collected 0 tests, so source contributors should expect investigation before productive work. Use the published runtime when direct Node access from a web interface is worth owning its security boundary and per-platform packaging.

Full review →
399

glide

bumptech/glide · ★ 35k1/7d · Java

Glide is an Android library that fetches, decodes, resizes, caches, and displays images, animated GIFs, and video stills. It is aimed at image-heavy screens where repeated downloads and decoding can make scrolling stutter, and it can use OkHttp or Volley instead of its default network layer.

SetupDocsCommunityMaturity

Verdict Our Glide build and test commands both failed after a 116-second install because Gradle could not find the Java 11 toolchain it requested. Use Glide v5 for an API 23 or newer Android app when its caching, decoding, and established View workflow fit better than a newer Kotlin-first loader.

Full review →
400

carbon-lang

carbon-language/carbon-lang · ★ 34k · C++

Carbon is an experimental programming language and toolchain for organizations with large C++ codebases. It tries to make new code easier to read and safer while letting teams call existing C++ and migrate in pieces, although its own README says it is not ready for use.

SetupDocsCommunityMaturity

Verdict Our Carbon checkout installed 35 packages in 16 seconds and passed its 4-second build step, but it exposed no test target for our sandbox to run. That is enough for language research, not a reason to place production work on an experimental toolchain whose own roadmap puts usable 1.0 beyond the 0.2 evaluation milestone.

Full review →
401

folly

facebook/folly · ★ 31k10/7d · C++

Folly is Meta's collection of C++20 building blocks for infrastructure and performance-sensitive applications. It fills gaps left by the standard library and Boost, especially where Meta has needed different behavior or a tighter performance profile at large scale.

SetupDocsCommunityMaturity

Verdict Our build completed in 7 seconds, but the test step ended with 2 collection errors and zero tests executed. Folly is a serious choice for experienced C++ infrastructure teams that need its specific components or inherit it through Meta's ecosystem, not a default utility dependency for every application.

Full review →
402

deskflow

deskflow/deskflow · ★ 29k · C++

Deskflow lets one computer's keyboard, mouse, or trackpad control other nearby computers. It solves the desk-clutter problem of switching input devices between machines, while also sharing the clipboard and encrypting connections with TLS by default. It is a software KVM without video switching.

SetupDocsCommunityMaturity

Verdict Our install attempt failed after 6 seconds because CMake could not find Qt 6.7.0. Deskflow is still the open-source software KVM I would try first for a mixed-computer desk, thanks to default TLS, clipboard sharing, Wayland support, broad platform coverage, and development activity as recent as September 10, 2026.

Full review →
403

beads

gastownhall/beads · ★ 27k200/7d · Go

Beads is a command-line issue tracker designed to give coding agents persistent tasks, dependencies, decisions, and handoff context inside a project. It stores the working graph in Dolt, so agents can ask what is ready, claim work, record history, and sync the database across clones without relying on a long Markdown checklist.

SetupDocsCommunityMaturity

Verdict Our Beads run installed 668 packages and built in 162 seconds, but 17 of 150 test groups failed after 464 seconds. Its dependency graph and agent integrations solve a real continuity problem, yet Dolt storage, schema coordination, and the current claim race make it an operational system rather than a fancy TODO file.

Full review →
404

cli

spicetify/cli · ★ 25k70/7d · JavaScript

Spicetify CLI customizes the official Spotify desktop client on Windows, macOS, and Linux. It lets people change the interface, inject CSS, add extensions, and install custom apps instead of accepting Spotify's stock desktop experience.

SetupDocsCommunityMaturity

Verdict Our install failed after 21 seconds, before any build or tests could run. Spicetify CLI is still the compelling choice when the goal is deep customization of the official Spotify desktop client, and its 24,380 stars, same-day push, and 13 open issues indicate an active, widely adopted project.

Full review →
405

v86

copy/v86 · ★ 23k19/7d · JavaScript

v86 is a JavaScript x86 PC emulator that translates machine code into WebAssembly at runtime, letting a browser boot older and 32-bit operating systems without a native virtual machine. It solves the awkward problem of putting an interactive legacy PC, teaching environment, or OS demo directly on a web page.

SetupDocsCommunityMaturity

Verdict Our install completed in 12 seconds, but our runner found no build or test target to exercise. v86 is an unusually capable choice when the browser itself must host a 32-bit x86 machine, and its demos plus embedding library make the value easy to see.

Full review →
406

waveterm

wavetermdev/waveterm · ★ 22k54/7d · Go

Wave Terminal is a cross-platform terminal that combines command-line sessions, remote access, file editing, previews, and an AI assistant in one workspace. It is meant to reduce the window switching involved in debugging locally or working on remote machines, while letting you use your own API keys or local models.

SetupDocsCommunityMaturity

Verdict All 46 tests passed in our run, but npm audit still found 65 known vulnerabilities. Wave is worth trying if durable SSH, inline previews, and bring-your-own-model AI solve real friction in your daily terminal work.

Full review →
407

tabler-icons

tabler/tabler-icons · ★ 22k63/7d · JavaScript

Tabler Icons is a large, MIT-licensed collection of SVG icons for websites and apps. It gives teams one consistent 24x24 visual system, then packages that system for direct SVG use, sprites, webfonts, popular JavaScript frameworks, and several export formats.

SetupDocsCommunityMaturity

Verdict Our build failed after 830 seconds because `rsvg-convert` was not found. That contributor experience needs better prerequisite documentation, and the 104-second test failure means we cannot call the checked-out monorepo clean on our box.

Full review →
408

gitbutler

gitbutlerapp/gitbutler · ★ 22k21/7d · Rust

GitButler is a source-available desktop app and command-line client for Git. It keeps parallel and stacked branches in one workspace, lets you reshape commits without an interactive rebase, and gives coding agents the same branch model through the `but` CLI.

SetupDocsCommunityMaturity

Verdict Our GitButler install failed after 9 seconds before it produced a build or test result, so source contributors should expect to fix the runtime and pnpm configuration before judging the code. The packaged client is still worth trying for developers who handle stacked branches and frequent commit surgery, because parallel branches and the operation history address real Git friction.

Full review →
409

CleanArchitecture

jasontaylordev/CleanArchitecture · ★ 21k21/7d · C#

CleanArchitecture is a .NET solution template for starting enterprise-style ASP.NET Core applications with clear boundaries between application layers. It generates an Angular, React, or API-only project and lets you choose SQLite, PostgreSQL, or SQL Server, saving teams from assembling that foundation by hand.

SetupDocsCommunityMaturity

Verdict Our build failed after 10 seconds, and the test command hit the 900-second cap. Use CleanArchitecture when your team already understands .NET layering and wants a credible, current starting point with real choices for UI and storage.

Full review →
410

bettercap

bettercap/bettercap · ★ 20k31/7d · Go

bettercap is a Go-based network security toolkit for examining and manipulating WiFi, Bluetooth Low Energy, Ethernet, CAN-bus, and related traffic. It gives security researchers one extensible console, API, and web interface for reconnaissance, packet inspection, spoofing, capture, and controlled attack work.

SetupDocsCommunityMaturity

Verdict Our build succeeded in 36 seconds, but 87 of 95 test packages failed in our sandbox. bettercap is compelling when an authorized operator genuinely needs one console spanning WiFi, BLE, CAN-bus, spoofing, sniffing, and automation.

Full review →
411

system-design-notes

liquidslr/system-design-notes · ★ 20k2.6k/7d

This is a set of English study notes based on volumes 1 and 2 of System Design Interview: An Insider's Guide. It turns the books' interview scenarios into a browsable chapter outline and adds links to deeper technical reading.

SetupDocsCommunityMaturity

Verdict We could not run install, build, or tests because the repository has no supported language ecosystem and no Dockerfile. Use it as a free companion index if you already own or are reading the two source books, especially when you want links from an interview prompt to primary engineering material.

Full review →
412

iii

iii-hq/iii · ★ 19k13/7d · Rust

iii is a shared runtime where services register callable functions and declare the triggers that invoke them. It tries to replace separate integration work for HTTP, queues, cron, state, streams, observability, sandboxes, and agents with one live catalogue, one routing model, and one operations console.

SetupDocsCommunityMaturity

Verdict Our iii run installed 820 packages, then produced 0 successful build tasks and repeated the same docs failure during the test command, so platform teams should trial the runtime from a pinned release before standardizing on it. The worker-function-trigger model is genuinely coherent for polyglot services and agent-discovered capabilities.

Full review →
413

fscan

shadow1ng/fscan · ★ 15k28/7d · Go

fscan is a Chinese-first internal-network scanner for authorized security testing; an English README and English interface mode are available. It combines host discovery, port and service identification, web fingerprints, weak-credential checks, vulnerability probes, and post-compromise modules in one Go binary.

SetupDocsCommunityMaturity

Verdict Our fscan run built in 47 seconds and passed all 21 tests in 136 seconds, so the checked Go code has a clean baseline. Use it only inside a documented assessment where its many active modules are an intentional choice and every finding is independently confirmed.

Full review →
414

bytebase

bytebase/bytebase · ★ 14k16/7d · Go

Bytebase is a control plane for database changes, access, and compliance. It gives developers, DBAs, security teams, and AI agents one governed place to request changes, review SQL, query data, and leave an audit trail instead of stitching together migration scripts, SQL clients, and tickets.

SetupDocsCommunityMaturity

Verdict Our build completed in 385 seconds, but 29 of 120 test packages failed, so Bytebase is credible software that still demands careful evaluation before production adoption. Use it when database governance is a real organizational problem and its combined review, access, masking, and audit features can replace several disconnected tools.

Full review →
415

gpui-kit

longbridge/gpui-kit · ★ 14k225/7d · Rust

GPUI Kit is a Rust toolkit for building desktop applications on macOS, Windows, and Linux. It supplies styled controls, lower-level interface behavior, and an optional JavaScript extension runtime, so a team does not have to assemble those pieces directly on top of GPUI.

SetupDocsCommunityMaturity

Verdict Our GPUI Kit run installed 1,018 packages in 45 seconds, then both the build and tests failed because the linker could not find `xkbcommon-x11`. Try version 0.6 if you are building a serious Rust desktop product and can own native dependencies, cross-platform testing, and API migration work.

Full review →
416

ipatool

majd/ipatool · ★ 11k239/7d · Go

ipatool is a cross-platform command-line program for searching the iOS and tvOS App Store, obtaining an app license, listing available versions, and downloading encrypted IPA packages. It gives developers and researchers a scriptable route to packages their Apple account already owns.

SetupDocsCommunityMaturity

Verdict Our ipatool run built in 46 seconds, but only 4 of 15 package test results passed before 11 build failures, so unattended use needs its own recovery checks. Use it for research with a separate Apple account and validation around every downloaded file.

Full review →
417

wloc

Yu9191/wloc · ★ 10k · JavaScript

wloc is a Chinese-language iOS testing tool that changes the coordinates returned by Apple's Wi-Fi and cell-based location service. The repository has no English setup guide. It combines proxy scripts, a map picker, and optional Cloudflare Worker hosting so an authorized tester can switch or clear a simulated network location without typing coordinates.

SetupDocsCommunityMaturity

Verdict Our wloc Worker install took 7 seconds and all 22 tests passed, but iOS 27 beta 6 blocks the certificate interception the complete tool depends on. It is a sensible, small utility for authorized testing on an older compatible device when your team reads Chinese and already understands its proxy app.

Full review →
418

windhawk

ramensoftware/windhawk · ★ 9.0k78/7d · Rust

Windhawk is a Windows customization platform that changes running programs through global injection and hooking. It gives users a marketplace for mods while keeping the desktop application, engine libraries, and interface code available for inspection and modification.

SetupDocsCommunityMaturity

Verdict Our Windhawk build failed after 142 seconds, and its test step hit the same 12 `windows-future` compiler errors, so this Debian checkout was not contributor-ready. Windows power users should still consider the packaged app when a specific trusted mod solves a stubborn interface problem.

Full review →
419

plannotator

backnotprop/plannotator · ★ 8.7k147/7d · TypeScript

Plannotator opens coding-agent plans, documents, HTML pages, and code diffs in a local browser so a person can mark them up and return structured feedback to the agent. It integrates with Claude Code, Codex, Copilot CLI, Gemini CLI, OpenCode, Kiro, Droid, Amp, and Pi.

SetupDocsCommunityMaturity

Verdict Our Plannotator install pulled 1,031 packages and 2,263 MB, then build and test both stopped because Bun was missing, so source contributors must treat the Bun toolchain as a hard prerequisite. The released binary is worth trying if agent plans keep getting approved in cramped terminal prompts and you want comments anchored to the actual text or diff.

Full review →
420

deepsec

vercel-labs/deepsec · ★ 8.0k40/7d · TypeScript

Deepsec is a self-hosted code-security scanner that combines fast pattern matching with coding agents that investigate suspected vulnerabilities. It is aimed at deep, on-demand reviews of existing large repositories, with resumable runs and optional distributed workers.

SetupDocsCommunityMaturity

Verdict Our Deepsec run built in 58 seconds and passed 2,379 of 2,411 tests, but one assertion failed, so the codebase is close to green rather than clean at commit a008d3e. Use it for a funded, human-reviewed security sweep where deep agent investigation may uncover flaws cheaper tools miss.

Full review →
421

agent-skills

tech-leads-club/agent-skills · ★ 6.3k · TypeScript

Agent Skills is a catalog, installer, and read-only MCP server for adding reusable instructions to coding agents. It solves the messy part of putting the same skills into Claude Code, Codex, Cursor, and other supported clients without copying folders by hand.

SetupDocsCommunityMaturity

Verdict Our Agent Skills run installed 1,603 packages and 1,024 MB, passed all 76 tests, and still returned 30 npm advisories, so this is a useful catalog to inspect rather than a security shortcut. Use it when several coding agents need the same curated skills and you can review each skill, license, and destination.

Full review →
422

tinycast

abue-ammar/tinycast · ★ 5.3k · Swift

Tinycast is a free, open-source command launcher for macOS that combines app search, clipboard history, shortcuts, notes, window controls, and optional AI commands in one native palette. It is aimed at people who like Raycast's workflow but want a SwiftUI and AppKit app with no telemetry and support for running many Raycast extensions.

SetupDocsCommunityMaturity

Verdict Our Tinycast sandbox stopped after 13 seconds because pnpm refused to run two esbuild build scripts, so evaluating the source is a deliberate macOS developer task rather than a one-command checkout. The released app is still an appealing choice for Mac users who want an open launcher with local features and can accept self-signing, Accessibility permission, and uneven Raycast compatibility.

Full review →
423

security-audit-skill

cloudflare/security-audit-skill · ★ 5.1k1.9k/7d · JavaScript

Cloudflare Security Audit Skill gives a coding agent a six-phase process for reviewing source code, testing suspected vulnerabilities, rejecting false positives, and producing human-readable and machine-readable findings. It solves a common agent-audit failure: generating a long checklist of theoretical weaknesses without proving that an attacker can exploit them.

SetupDocsCommunityMaturity

Verdict Use this skill when you already trust your coding agent with a repository and want a disciplined, adversarial workflow for finding new bugs that scanners may miss. Its strongest idea is separation of duties: hunters propose findings, fresh agents try to kill them, and another pass checks the final report against source.

Full review →
424

ligolo-ng

nicocha30/ligolo-ng · ★ 5.0k · Go

Ligolo-ng gives penetration testers routed access to a remote network through an agent that calls back over TLS. It creates a TUN interface on the operator's machine, so familiar tools can reach the remote network without sending their traffic through proxychains.

SetupDocsCommunityMaturity

Verdict Our Ligolo-ng run installed 198 packages and built successfully, but 3 of 9 test targets failed, so pentesters should trial the exact routing and scanning workflow before adopting it. The TUN design is easier to work with than per-tool SOCKS settings, and the unprivileged agent is a practical advantage.

Full review →
425

cli-printing-press

mvanhorn/cli-printing-press · ★ 4.7k47/7d · Go

CLI Printing Press researches an API or website, then generates a Go command-line client, a matching MCP server, local SQLite storage, agent instructions, and verification artifacts. It aims to produce task-oriented commands rather than stopping at one wrapper command per endpoint.

SetupDocsCommunityMaturity

Verdict Our CLI Printing Press build passed after 176 seconds, but the 614-second test run finished with 22 packages passing and 2 failing. Its factory can save serious work for teams that repeatedly build agent-facing API tools, provided every live probe and generated artifact is reviewed in isolation.

Full review →
426

atlas

pacifio/atlas · ★ 4.6k1.3k/7d · Rust

Atlas is a macOS desktop workspace that runs several coding agents against the same repository and records which session produced each Git commit. It keeps prompts, tool calls, file changes, notes, and local memory together so a developer can switch between Claude Code, Codex, Atlas's own agent, or another ACP agent without losing the project thread.

SetupDocsCommunityMaturity

Verdict Our Atlas build completed in 55 seconds and its tests passed in 14 seconds after a 958-package, 1,271 MB install. It is a credible trial for a macOS developer who uses more than one coding agent and wants commits tied back to the sessions that made them.

Full review →
427

claude-plugins-community

anthropics/claude-plugins-community · ★ 4.1k401/7d · Python

Claude Plugins Community is Anthropic's read-only mirror of community plugins approved for Claude Code and Claude Cowork. It gives users one installable marketplace and gives plugin authors a distribution route, while the actual submissions and reviews happen outside this repository.

SetupDocsCommunityMaturity

Verdict Use this marketplace when you want broad Claude plugin choice and are prepared to inspect each plugin before installation. The pinned sources and Anthropic review pipeline are useful filters, yet current reports of stale or broken entries make local due diligence necessary.

Full review →
428

dsh-anchored-standard

xiaobright/dsh-anchored-standard · ★ 3.8k1/7d · JavaScript

dsh-anchored-standard is an experimental set of DeepSeek Harness presets that changes the tools and context shown on a session's first request, then opens more tools after the initial response. Its goal is to preserve a preferred DeepSeek reasoning style without keeping the agent on the Minimal preset for the whole task; the main README is English and a Chinese version is linked.

SetupDocsCommunityMaturity

Verdict Our dsh-anchored-standard run needed 0 installed packages and passed all 205 tests in 9 seconds, but its current generic preset has no matching 98/99 benchmark. Use it as a research kit for DeepSeek Harness request shaping, with pinned host code and exported-header checks.

Full review →
429

DSH-better-sidebar

omdsh-dev/DSH-better-sidebar · ★ 3.6k155/7d · TypeScript

DSH Better Sidebar is a Chinese-first workbench plugin for DeepSeek Harness, and a maintained English README is available. It adds file browsing and editing, a real terminal, Git controls, previews, background-task views, floating panels, and side conversations to DSH while letting other plugins register their own tabs and file viewers.

SetupDocsCommunityMaturity

Verdict Our DSH Better Sidebar install stopped after 8 seconds inside pnpm, so the source checkout never reached a build or test on our box. Existing DSH users may still find it worth the setup because files, terminal, Git, tasks, and plugin tabs live beside the conversation.

Full review →
430

airgorah

martin-olivier/airgorah · ★ 3.5k437/7d · Rust

Airgorah is a Linux desktop application for auditing WiFi networks you own or are authorized to test. It puts traffic capture, client discovery, deauthentication, handshake capture, and access-point password cracking behind a GTK4 interface built around the aircrack-ng tool suite.

SetupDocsCommunityMaturity

Verdict Our build completed in 86 seconds, and all 48 tests passed with 0 failures. Airgorah is a credible choice if you already have a Linux wireless lab and want aircrack-ng workflows in a focused GTK4 interface.

Full review →
431

lazycodex

code-yeongyu/lazycodex · ★ 3.5k32/7d · TypeScript

LazyCodex packages the OmO agent harness for Codex, adding project memory, planning, persistent execution, specialist agents, hooks, and completion checks. It is for developers who want a strongly opinionated workflow layered over Codex instead of assembling those pieces themselves.

SetupDocsCommunityMaturity

Verdict Our LazyCodex run installed 0 packages in 5 seconds and passed all 9 tests in 8 seconds, so its repository checks are light despite a 215.9 MB checkout. Use it if native Codex feels too loose and you are prepared to audit an opinionated layer after each upgrade.

Full review →
432

iloader

nab138/iloader · ★ 3.4k · TypeScript

iloader is a desktop utility for installing SideStore, LiveContainer, and other IPA files on an iPhone or iPad from a computer. It also handles pairing files, development certificates, and app IDs, reducing the manual work around Apple sideloading.

SetupDocsCommunityMaturity

Verdict Our build completed in 11 seconds after a 42-second install. iloader looks like a practical front end for people already committed to SideStore or IPA sideloading, especially because it combines installation, pairing-file placement, and certificate cleanup.

Full review →
433

argent

software-mansion/argent · ★ 2.8k165/7d · TypeScript

Argent is an MCP toolkit that lets coding agents control, inspect, record, and profile apps on iOS simulators, Android devices, TVs, Electron, and Chromium. It gives Claude Code and other supported agents device actions such as tapping, taking screenshots, reading UI trees, replaying flows, and collecting diagnostics.

SetupDocsCommunityMaturity

Verdict Our Argent run installed 501 packages, built in 30 seconds, and exposed 11 known vulnerabilities while providing no test target, so it needs a security and behavior review before an agent touches a real app. Its device breadth is useful for an experienced mobile team that wants interactive diagnosis inside Claude Code or another MCP client.

Full review →
434

mubeng

mubeng/mubeng · ★ 2.7k157/7d · Go

Mubeng checks lists of HTTP and SOCKS proxies, then runs a local HTTP proxy that rotates outgoing requests across the working entries. It is aimed at developers and security testers who already have a proxy pool and want one small command-line tool to validate and use it.

SetupDocsCommunityMaturity

Verdict Our Mubeng run built in 47 seconds and passed all 3 tests in 17 seconds, so trying the 1.2 MB checkout is cheap. Use it for an authorized proxy pool when mixed upstream protocols and simple HTTP rotation fit the job.

Full review →
435

bilibili-video-downloader

lanyeeee/bilibili-video-downloader · ★ 2.0k36/7d · Rust

Bilibili Video Downloader is a Chinese-language desktop app for saving Bilibili videos, series, courses, subtitles, danmaku, covers, and media-library metadata. The README and issue tracker are primarily Chinese, and the repository does not provide an English README.

SetupDocsCommunityMaturity

Verdict Our run installed 385 packages in 18 seconds and built in 28 seconds, but the project supplied no automated test target for its 22,766 lines of source. Chinese-reading users who want Bilibili downloads prepared for Emby get a thoughtful mix of GUI task management, subtitles, danmaku, chapters, and NFO output.

Full review →
436

OpenCADStudio

HakanSeven12/OpenCADStudio · ★ 2.0k666/7d · Rust

Open CAD Studio is a Rust desktop and browser application for editing 2D drawings and building 3D solids. It reads and writes DWG and DXF files directly, giving engineers and drafters an open-source route that does not depend on a conversion service.

SetupDocsCommunityMaturity

Verdict Our Open CAD Studio build took 459 seconds and succeeded, but its tests stopped after 71 seconds because `mesh_from_solid` was missing during test compilation. Try it for accessible DWG and DXF work, especially if the browser editor or JSON automation solves a specific problem.

Full review →
437

nodeterm

eneskirca/nodeterm · ★ 1.9k62/7d · TypeScript

nodeterm is a desktop and browser workspace that places live terminals and coding-agent sessions on a zoomable canvas instead of hiding them behind tabs. It uses tmux to keep sessions running, adds agent status and a kanban view, and helps developers track several local or remote jobs.

SetupDocsCommunityMaturity

Verdict Our nodeterm build passed in 38 seconds, but 36 of 8,973 tests failed, so the packaged app is easier to recommend for a trial than the current checkout is for contribution. Use it when spatial layout genuinely helps you supervise several agents and branches.

Full review →
438

context-engineering-kit

NeoLabHQ/context-engineering-kit · ★ 1.7k21/7d · TypeScript

Context Engineering Kit is a GPL-licensed collection of skills, commands, agents, and hooks for coding assistants, with Claude Code as its fullest installation target. It packages workflows for reflection, specifications, reviews, testing, documentation, Git, architecture rules, and MCP setup so teams do not have to write each instruction set themselves.

SetupDocsCommunityMaturity

Verdict Our Reflexion hook install added 59 packages in 35 seconds, then all 41 tests passed in 7 seconds, so that executable slice is easy to trial and has a working check. Use Context Engineering Kit when Claude Code is the main client and you want selectable, inspectable workflows for review or specification work.

Full review →
439

zero

Gitlawb/zero · ★ 1.7k6/7d · Go

Zero is a local terminal coding agent with a Go core and an npm-distributed launcher. It can inspect and edit repositories, run commands, use multiple model providers, keep sessions on disk, and expose its tools through interactive or machine-readable interfaces.

SetupDocsCommunityMaturity

Verdict Our npm-path run installed 83 packages in 10 seconds, exposed no build or test target, and reported 2 moderate vulnerabilities. Zero is worth trying if model choice, local sessions, and explicit permissions outweigh the cost of adopting a fast-moving v0.8 project.

Full review →
440

gortex

zzet/gortex · ★ 1.6k17/7d · Go

Gortex indexes source code into a local graph so coding agents and developers can ask for symbols, usages, call chains, contracts, and likely change impact without reading whole files. It ships as a CLI, daemon, web interface, HTTP API, and MCP server, with support for multiple repositories in one workspace.

SetupDocsCommunityMaturity

Verdict Our Gortex run spent 271 seconds building and 596 seconds testing before 80 of 81 test groups passed, so this is a substantial code-intelligence system with one unresolved lab failure. Trial it when local cross-repository graph queries could save agents repeated file reads, but verify search counts and memory on a copy of your own estate.

Full review →
441

traycer

traycerai/traycer · ★ 1.5k89/7d · TypeScript

Traycer is a desktop workspace for running Claude Code, Codex, Cursor, OpenCode, and Traycer agents beside shared tasks, terminals, code, and conversations. It keeps agent sessions durable, lets several agents communicate or work in parallel, and adds team boards and cross-device state around provider subscriptions you already have.

SetupDocsCommunityMaturity

Verdict Our Traycer run installed 1,828 packages and built in 59 seconds, but its test command was still running when our 900-second cap ended it despite thousands of passing assertions already printed. Traycer is worth trying if agent sprawl has become a desktop coordination problem and cloud-backed collaboration is acceptable.

Full review →
442

luvus

RizRiyz/luvus · ★ 84173/7d · Rust

Luvus is a terminal workspace for running and supervising several AI coding-agent sessions. It keeps panes and projects alive, shows agent status and usage, manages Git worktrees, and exposes the same workspace through a desktop terminal, SSH, or a narrow-screen client.

SetupDocsCommunityMaturity

Verdict Our Luvus build passed in 77 seconds and 1,838 of 1,840 tests passed, but the 2 failures make copy-mode and viewport behavior worth retesting before daily use. It is a good fit for one developer juggling enough agent sessions that tmux titles and separate worktrees have become hard to track.

Full review →
443

common-skills

warpdotdev/common-skills · ★ 57718/7d · Python

Common Skills is Warp's public collection of reusable instructions for coding agents, covering specifications, pull requests, debugging, research, reviews, and skill maintenance. It lets teams copy selected workflows into a project or install them globally through the `skills` CLI.

SetupDocsCommunityMaturity

Verdict Our sandbox could not run commit f589e22 because it has no supported package ecosystem or Dockerfile, so Common Skills arrives without lab evidence for installation or tests. The collection is worth reading and selectively copying, especially for spec and review workflows, but a global `*` install imports too many Warp-specific assumptions for most outside teams.

Full review →
444

Lighthouse

HarbourMasters/Lighthouse · ★ 4308/7d · C

Lighthouse is a native PC port of Banjo-Kazooie that creates its game data from a retail ROM supplied by the player. It adds modern display, controller, language-pack, romhack, and custom-asset options without distributing Nintendo's copyrighted game files.

SetupDocsCommunityMaturity

Verdict Our Lighthouse checkout installed in 15 seconds and built in 5 seconds, but pytest collected 0 tests, so the source passed compilation without giving us an automated gameplay safety net. Use the packaged release if you own a supported ROM and want Banjo-Kazooie with modern controls, display options, and romhack support.

Full review →
445

nippo

nwiizo/nippo · ★ 1772/7d · Rust

Nippo is a Japanese-language Rust CLI and companion skill that turns local Claude Code, Codex, and opencode session history into daily reports and work reflections. Its documentation and generated reports are Japanese, and the repository does not include an English README. The collector handles counts and dates while the installed agent skill writes the narrative.

SetupDocsCommunityMaturity

Verdict Our Nippo run installed 55 packages in 16 seconds and passed all 106 tests, so it earns a trial for Japanese-speaking developers already living in supported coding agents. Use it when local transcripts are the missing raw material for a personal review habit.

Full review →
446

Python

TheAlgorithms/Python · ★ 225k221/7d · Python

TheAlgorithms/Python is a large, browsable collection of algorithms implemented in Python for education. It gives learners and contributors readable examples to study, compare, and improve, rather than a production library to install behind an application.

SetupDocsCommunityMaturity

Verdict Our build completed in 5 seconds, but the test run still ended with 3 failures and 186 collection or setup errors. Use TheAlgorithms/Python as a generous study shelf and contribution playground, not as a dependency whose implementations you trust by default.

Full review →
447

30-Days-Of-Python

Asabeneh/30-Days-Of-Python · ★ 74k706/7d · Python

30 Days of Python is a free, English-language course that takes a beginner from installing Python through data work, web APIs, Flask, and MongoDB. Each numbered day combines an explanation, examples, and exercises, with translation folders for several languages. It solves the blank-page problem for self-directed learners by supplying an ordered route through a large subject.

SetupDocsCommunityMaturity

Verdict Our `python_for_web/` install pulled 41 packages and pip-audit found 19 known vulnerabilities, so use this course as reading material rather than deployable starter code. The 30-day sequence is friendly and unusually broad for a free beginner resource, but old version guidance and unverified examples require an alert learner or instructor.

Full review →
448

union

unionlabs/union · ★ 74k · Rust

Union is a cross-chain messaging and asset-transfer protocol that uses light clients and zero-knowledge proofs instead of a trusted multisignature bridge. This repository contains the network node, prover, relayer, smart contracts, TypeScript SDKs, web applications, and deployment machinery needed to build and operate that system.

SetupDocsCommunityMaturity

Verdict Union is serious protocol infrastructure for teams that specifically want light-client-based interoperability and can evaluate its chain-specific code. It is a poor casual dependency: the monorepo, Nix build, prover, relayer database, contracts, and node operations demand specialist ownership.

Full review →
449

omarchy

omacom/omarchy · ★ 41k1.9k/7d · Shell

Omarchy is an opinionated Linux distribution that packages a polished desktop, developer tools, applications, themes, and operating conventions into one system. It aims to spare developers from assembling and documenting an Arch-based workstation piece by piece.

SetupDocsCommunityMaturity

Verdict We could not run Omarchy at commit 56fbaf4 because our Shell runner had no supported ecosystem and the repository had no Dockerfile. Treat it as a whole-workstation choice, not a package you can casually trial inside an application container.

Full review →
450

v

vlang/v · ★ 38k43/7d · V

V is a compiled programming language that aims to feel small and readable while producing native programs through C and other backends. It combines a compiler, package manager, formatter, documentation tool, web framework, database support, graphics libraries, and cross-compilation tools in one repository.

SetupDocsCommunityMaturity

Verdict Our sandbox did not run commit 4ddbd6c because V is outside the lab's supported ecosystems, so this review cannot confirm the README's quick source-build story. V 0.5.2 is worth a personal trial for developers drawn to its small syntax, C output, and unusually broad standard tooling.

Full review →
451

shadPS4

shadps4-emu/shadPS4 · ★ 33k97/7d · C++

shadPS4 is an early PlayStation 4 emulator core for Windows, Linux, and recent Apple-silicon Macs. It lets owners run supported games from their own dumps, but end users are directed to a separate Qt launcher and must check compatibility title by title.

SetupDocsCommunityMaturity

Verdict Our shadPS4 run installed 151 packages in a nested documentation project and found 49 npm vulnerabilities, but it did not build or test the C++ emulator. End users should try the official QtLauncher only after checking their exact game in the compatibility tracker and accepting that current releases still regress.

Full review →
452

Recordly

webadderallorg/Recordly · ★ 29k2.9k/7d · TypeScript

Recordly is an open-source desktop screen recorder and editor for product demos, walkthroughs, and short instructional videos. It records the screen, microphone, system audio, and webcam, then adds zooms, cursor treatment, annotations, framing, and MP4 or GIF export in the same app.

SetupDocsCommunityMaturity

Verdict Our Recordly run passed 1,083 tests, but its Linux build stopped after 83 seconds for a missing `GH_TOKEN`, and the install occupied 1,209 MB with 27 known vulnerabilities. It is worth trying for short, styled product demos on a supported macOS or Windows machine, where its recorder-to-editor flow can replace several manual steps.

Full review →
453

Wand-Enhancer

k1tbyte/Wand-Enhancer · ★ 27k1.5k/7d · C#

WandEnhancer is a Windows patcher for the Wand (WeMod) desktop app. It changes the local interface, adds custom client scripts, and provides a phone-friendly control panel for an active trainer, solving problems the official client does not expose as settings.

SetupDocsCommunityMaturity

Verdict Our WandEnhancer run installed 397 packages and passed all 34 web-panel tests, but it did not prove the Windows patcher against Wand itself. Use it if the custom client and phone controls justify building an unsigned artifact and repatching after compatibility changes.

Full review →
454

fhevm

zama-ai/fhevm · ★ 25k · Rust

FHEVM is Zama's framework for running smart-contract logic over encrypted values on EVM-compatible blockchains. Solidity contracts record symbolic operations on the host chain, while an off-chain coprocessor performs the encrypted computation and a key-management system controls decryption.

SetupDocsCommunityMaturity

Verdict Our FHEVM install pulled 2,587 packages and 2,071 MB, then its test command failed in 5 seconds because Bun was absent and npm audit reported 108 known vulnerabilities. That is enough friction to reject it for a casual private-token experiment, especially when commercial use needs a separate patent license.

Full review →
455

happy

slopus/happy · ★ 24k · TypeScript

Happy is an open-source mobile, web, and macOS client for controlling Claude Code or Codex sessions running on your computer. It solves the practical problem of checking progress, answering permission prompts, and switching control between your desk and another device without sending code across the network unencrypted.

SetupDocsCommunityMaturity

Verdict Our source install failed after 8 seconds with a pnpm configuration conflict across 9 workspaces. Happy is still one of the more coherent answers to controlling Claude Code or Codex from a phone, with encrypted sync, push alerts, and clients for the major personal-device platforms.

Full review →
456

core

doomemacs/core · ★ 23k20/7d · Emacs Lisp

Doom Emacs is a configuration framework that turns GNU Emacs into an opinionated editor with Vim-style keys, language tools, workspaces, and curated packages. It reduces the work of assembling a personal Emacs setup while leaving your private configuration and system dependencies in your hands.

SetupDocsCommunityMaturity

Verdict We could not run commit 2e0dbad because our harness does not support Emacs Lisp and the repo has no Dockerfile, so there is no lab result to hide behind. Doom Emacs is worth trying for an experienced Emacs or Vim user who accepts its sync commands, external dependencies, and moving master branch.

Full review →
457

RustPython

RustPython/RustPython · ★ 22k5/7d · Rust

RustPython is a Python 3 interpreter written in Rust rather than a wrapper around CPython. It is useful when a Rust application needs embedded Python, or when Python code must run in WebAssembly or WASI without shipping the usual CPython runtime.

SetupDocsCommunityMaturity

Verdict Our RustPython build and 2-test Cargo suite both passed, but they consumed 947 seconds combined and do not establish CPython 3.14 compatibility across a 1,311,696-line checkout. Try it for Rust embedding or WebAssembly, where its architecture solves a problem CPython does not solve neatly.

Full review →
458

no-mistakes

kunchenguid/no-mistakes · ★ 8.5k100/7d · Go

No Mistakes is a local Git gate that checks work in a disposable worktree before it reaches the real remote. It uses a configured coding agent to review, test, document, lint, repair, push, open a pull request, and watch CI, while asking a person to decide changes that may affect intent.

SetupDocsCommunityMaturity

Verdict Our No Mistakes run built in 41 seconds but finished with 16 of 17 tests passing, so this Git gate is not ready to become the only authority between a branch and its remote. Its isolated workflow is useful for supervised agent review, especially for solo developers already using Claude Code or Codex.

Full review →
459

Acode

Acode-Foundation/Acode · ★ 7.0k144/7d · JavaScript

Acode is an Android code editor for opening local files, editing several programming languages, and previewing websites on a phone or tablet. It adds a JavaScript console, SFTP access, SSH and an Alpine terminal, plus its own plugin system.

SetupDocsCommunityMaturity

Verdict Our Acode checkout passed all 377 tests in 36 seconds, but the APK build failed after 20 seconds because Cordova had no Android platform added. Installing the finished app from Google Play or F-Droid is the sensible route for users; source contributors should use the documented container or complete the Android toolchain setup first.

Full review →
460

plugins

openai/plugins · ★ 6.8k999/7d · JavaScript

OpenAI Plugins is a curated source collection for extending Codex with skills, app connections, MCP configuration, commands, agents, hooks, and supporting assets. It is mainly a reference and distribution repository for plugin authors and Codex users, not one application that you install and run as a service.

SetupDocsCommunityMaturity

Verdict Our plugin-eval install added 0 packages in 4 seconds, but its test command failed after 8 seconds on a literal glob path, so this repository is better used as pinned source material than as a trusted monolithic package. Codex plugin authors should study it because 53 manifests show the format in real bundles, and users can take only the plugins they need.

Full review →
461

armorpaint

armory3d/armorpaint · ★ 5.1k764/7d · C

ArmorPaint is a desktop application for painting physically based textures directly on 3D models. It combines brushes, layers, material nodes, baking, mesh tools, and local neural nodes so an artist can move from an unwrapped model to engine-ready texture files in one program.

SetupDocsCommunityMaturity

Verdict We could not run ArmorPaint commit 0a125e5 because our harness does not support C projects and the repository has no Dockerfile, so buyers should treat a source build as a hands-on evaluation. The feature set is unusually broad for an open-source texture painter, especially for PBR layers, baking, export presets, and local neural tools.

Full review →
462

ESP32-Cheap-Yellow-Display

witnessmenow/ESP32-Cheap-Yellow-Display · ★ 4.4k30/7d · Rust

ESP32-Cheap-Yellow-Display is a community documentation and example repository for the ESP32-2432S028 board, a small ESP32 module with a built-in color screen and resistive touch. It fills the gap left by inconsistent seller documentation with pin maps, setup notes, sketches, cases, projects, and troubleshooting advice.

SetupDocsCommunityMaturity

Verdict Our Rust example stopped after 7 seconds because the required custom `esp` toolchain was absent, so Rust users must complete Espressif-specific setup before they can even compile. The repository is still a useful field guide for the exact ESP32-2432S028 board, especially through Arduino or ESPHome examples.

Full review →
463

renodx

clshortfuse/renodx · ★ 4.2k300/7d · HLSL

RenoDX is a DirectX game-modding toolkit built on ReShade's add-on system. It lets mod authors replace shaders, change rendering resources, add controls, and inspect live frames so they can fix HDR output and other rendering behavior without patching each game executable.

SetupDocsCommunityMaturity

Verdict Our RenoDX web run installed 0 packages, built in 9 seconds, and had no test target, so its catalogue is easy to build but supplies no web regression gate. Use RenoDX when a supported game's rendering needs more than a post-processing filter and you are comfortable testing every update.

Full review →
464

Impactor

claration/Impactor · ★ 3.2k45/7d · Rust

Impactor is a cross-platform desktop app and command-line signer for installing IPA files on iPhones, iPads, Apple TV devices, and Macs. It uses an Apple account to create certificates and provisioning profiles, signs the app, talks to the device, and can also prepare pairing files for companion tools.

SetupDocsCommunityMaturity

Verdict Our Impactor checkout built in 360 seconds, but `cargo test` found 0 tests, so the green command is not evidence that signing and pairing work. Use the GUI for supervised sideloading when Xcode is unavailable, then inspect the IPA for a signature and provisioning profile before installation.

Full review →
465

AndroidMic

teamclouday/AndroidMic · ★ 1.7k · Rust

AndroidMic turns an Android phone into a microphone source for a Windows, Linux, or macOS computer. A phone app sends microphone audio over WiFi or USB to a desktop receiver, which plays it into a separate virtual audio device that calling, recording, and streaming software can select.

SetupDocsCommunityMaturity

Verdict Our AndroidMic build failed after 338 seconds, and its 47-second test command failed at the same missing PipeWire system library after Cargo installed 750 packages. Use the packaged 2.2.9 apps when you need an inexpensive spare microphone and can tolerate virtual-device setup.

Full review →
466

anti-mage

N4darae/anti-mage · ★ 1.7k111/7d · Go

Anti-Mage is a Go library and loopback web app that checks whether a browser's reported capabilities agree with one another. It returns a 0 to 100 score and a plain determination, but it does not identify a person, vendor, or specific automation tool.

SetupDocsCommunityMaturity

Verdict Our Anti-Mage run installed 1 package and passed all 10 tests, so trying its local Go scanner costs little. Use it as one browser-consistency signal or as an in-process building block, especially when false accusations matter.

Full review →
467

ai-toolbox

coulsontl/ai-toolbox · ★ 1.6k217/7d · Rust

AI Toolbox is a cross-platform desktop control panel for provider, model, prompt, session, MCP, skill, and gateway settings across coding assistants such as Claude Code, Codex, Gemini CLI, and OpenCode. Its main README and supporting docs are primarily Chinese, with no separate English README, although the application lists Chinese and English interface options. It replaces scattered configuration files and manual copying with one local database and synchronization layer.

SetupDocsCommunityMaturity

Verdict Our AI Toolbox run installed 785 packages and used 818 MB, then passed its 71-second build and 21-second test step. Use it if one technical user actively manages several coding agents and can audit the configuration changes it writes.

Full review →
468

DevOps-Interview-Guide

litu54/DevOps-Interview-Guide · ★ 1.5k21/7d

DevOps Interview Guide is an English-language collection of questions that contributors say they encountered in DevOps, SRE, cloud, and platform interviews. It groups each reported interview by company and role, helping candidates see the kinds of operational scenarios employers ask about without pretending to be a full study course.

SetupDocsCommunityMaturity

Verdict Use DevOps Interview Guide as a stack of flash cards, especially when your target company appears in the folders. Do not treat it as an authoritative syllabus or copy its questions into a commercial product without resolving the missing license.

Full review →
469

navop

feigeCode/navop · ★ 1.4k100/7d · Rust

Navop is a native desktop workspace that combines database clients, SSH and SFTP, terminals, remote desktops, monitoring, notes, and AI agents. Its main README is in English, with a Chinese translation, and it is aimed at people who would rather manage those connections in one application than keep several specialist tools open.

SetupDocsCommunityMaturity

Verdict Our Navop run installed 1,524 packages in 96 seconds, but its build hit the 900-second cap and its tests stopped on a missing `xkbcommon-x11` linker library. Try the packaged application if one native workspace would replace separate database, terminal, file transfer, and remote desktop clients.

Full review →
470

autoprompt-skill

Spielewoy/autoprompt-skill · ★ 1.2k122/7d · JavaScript

Autoprompt Skill installs a structured multi-agent coding workflow into nine supported coding assistants, including Claude Code and Codex. One explicit goal becomes a loop of planning, implementation, testing, independent review, repair, and final verification, with controls for concurrency and model routing.

SetupDocsCommunityMaturity

Verdict Our AutoPrompt test command finished with 78 of 80 tests passing and still exited 1, so version 1.0.4 is not a clean lab pass. Its explicit workflow is attractive for substantial coding jobs where independent review can justify more agent work.

Full review →
471

power-platform-skills

microsoft/power-platform-skills · ★ 87425/7d · JavaScript

Power Platform Skills is Microsoft's plugin marketplace for building Power Pages sites, Power Apps, Power Automate flows, MCP widgets, and related mobile projects with Claude Code or GitHub Copilot. It packages instructions, commands, agents, templates, and MCP connections around Microsoft's existing developer tools rather than replacing those tools.

SetupDocsCommunityMaturity

Verdict Our mobile-template install pulled 969 packages and 1,268 MB, while npm audit found 8 moderate vulnerabilities and the package exposed no generic build or test target. The marketplace is worth a controlled trial for experienced Power Platform teams because its 8 plugins encode many product-specific steps, but each plugin needs its own permission, telemetry, and verification review.

Full review →
472

hello-sdd

mouredev/hello-sdd · ★ 56065/7d · Python

Hello SDD is a Spanish-language course repository that teaches spec-driven development by preserving the prompts, specifications, plan, tasks, code, and tests for one small Python project. It does not provide an English README or English course notes. The useful idea is to inspect a complete paper trail for agent-assisted work instead of starting from an empty template.

SetupDocsCommunityMaturity

Verdict Our hello-sdd run installed 36 packages in 19 seconds and passed all 36 tests, so the worked example is unusually easy to verify. Use it as a Spanish teaching case for how a spec becomes tasks and tests, especially if you use Claude Code or OpenCode.

Full review →
473

proliferate

proliferate-ai/proliferate · ★ 50238/7d · TypeScript

Proliferate is an open-source workspace for running Claude Code, Codex, OpenCode, Cursor, Grok, and other coding agents side by side. Each task gets its own Git worktree, branch, terminal, conversation, and review state, while shared MCP connections, skills, browser tools, and recurring workflows coordinate work across agents.

SetupDocsCommunityMaturity

Verdict Our Proliferate install pulled 1,108 packages and used 1,128 MB, while the root offered no generic build or test target for our harness to run. Try it when parallel agents are already creating branch, worktree, credential, and review-state confusion.

Full review →
474

eigenwise-toolshed

Eigenwise/eigenwise-toolshed · ★ 280154/7d · JavaScript

Eigenwise Toolshed is a marketplace of six independent Claude Code plugins for repository maps, conditional rules, tracked parallel work, extra subscription models, local usage records, and guided setup. It is meant to stop each new coding session from rediscovering the project or rebuilding the same operating habits.

SetupDocsCommunityMaturity

Verdict Our Node 22 run installed Model Gateway's 6 packages in 26 seconds and used 3 MB, but 5 of 281 tests failed, so Toolshed deserves a plugin-by-plugin trial instead of blanket adoption. Codebase Mapper and Live Rules solve narrow problems with understandable files, while Sidequest and Model Gateway reach much further into work ownership and process routing.

Full review →
475

confetti

rhevorn/confetti · ★ 1711/7d · TypeScript

Confetti is a VS Code extension that detects, highlights, and formats more than 16 kinds of configuration files, including Nginx, dotenv, TOML, SSH, hosts, and crontab. It gives mixed repositories one local extension for files that would otherwise need several narrow extensions or fall back to plain text.

SetupDocsCommunityMaturity

Verdict Our Confetti run installed 496 packages, built in 14 seconds, and passed all 299 tests, so the code earns a trial for VS Code users juggling mixed configuration files. Its local operation and cautious detection are sensible, while the three-day public history and missing CI workflow leave maintenance quality unproven.

Full review →
476

andrej-karpathy-skills

multica-ai/andrej-karpathy-skills · ★ 213k1.3k/7d

andrej-karpathy-skills is a small set of behavioral instructions for Claude Code and Cursor. It asks a coding agent to expose assumptions, prefer simple code, limit edits to the requested scope, and work toward checks that can prove a task is done.

SetupDocsCommunityMaturity

Verdict Use andrej-karpathy-skills if your main complaint is that Claude Code guesses too much, edits too broadly, or turns small jobs into large designs. The rules are short enough to audit and adapt, but the packaging needs cleanup and the repository does not yet provide merged evaluation evidence.

Full review →
477

manim

3b1b/manim · ★ 94k290/7d · Python

ManimGL is the Python animation engine Grant Sanderson uses to make precise mathematical explanations for 3Blue1Brown videos. You describe equations, shapes, transformations, camera moves, and timing in code, which makes repeatable visuals easier than animating each change by hand.

SetupDocsCommunityMaturity

Verdict Our ManimGL run installed 33 packages and built in 2 seconds, but pytest stopped with 2 collection errors because `numpy` was missing. Choose it when you specifically want the engine and interactive habits used for 3Blue1Brown production.

Full review →
478

localstack

localstack/localstack · ★ 65k5/7d · Python

LocalStack runs an AWS-like environment in a local container so applications, infrastructure code, and CI jobs can call cloud APIs without creating real AWS resources. This repository contains the former open-source runtime, but it is now archived as development moves to LocalStack's unified image.

SetupDocsCommunityMaturity

Verdict LocalStack remains useful for local AWS integration work, but this archived repository is no longer the product I would adopt. New evaluations should use the unified image, check the exact service coverage and plan terms, and run contract tests against AWS before release.

Full review →
479

cool-retro-term

Swordfish90/cool-retro-term · ★ 26k · QML

cool-retro-term is a terminal emulator for Linux and macOS that makes a live shell look like an old cathode-ray tube display. Colors, fonts, screen curvature, glow, and other effects are adjustable from its settings. The shell underneath stays familiar; the visual treatment is the reason to install it.

SetupDocsCommunityMaturity

Verdict Our cool-retro-term sandbox installed 35 packages in 14 seconds and built in 5 seconds, but the checkout supplied no test target. Use a packaged build if you want its CRT effects, because the Qt6 application and Qt5 build wiki currently disagree.

Full review →
480

Seelen-UI

eythaann/Seelen-UI · ★ 18k56/7d · Rust

Seelen UI is a replacement desktop environment for Windows 10 and 11 that can take over the taskbar, dock, launcher, window tiling, flyouts, and virtual desktops. It is for people who want those parts configured in one application instead of assembling several Windows utilities.

SetupDocsCommunityMaturity

Verdict Our Node 22 sandbox could not finish Seelen UI's npm install in 54 seconds because the preinstall step called a missing Deno executable, so source contributors need a wider toolchain than the README states. Windows 10 and 11 users should try the signed release if they want one app to coordinate tiling, widgets, a dock, and shell styling.

Full review →
481

cockpit-tools

jlcodes99/cockpit-tools · ★ 18k381/7d · Rust

Cockpit Tools is primarily documented in Chinese, and a substantial English README exists. It is a desktop account switcher and quota dashboard for Codex, GitHub Copilot, Cursor, Windsurf, Kiro, Antigravity, Grok CLI, and several other AI coding products, with isolated instances and local gateway features.

SetupDocsCommunityMaturity

Verdict Our Cockpit Tools build succeeded in 77 seconds after a 198-package install, but the repository offered no general test target for us to run. The app can save real time for one technical user juggling many coding accounts, provided local credential files and a fast-changing release line are acceptable.

Full review →
482

microsandbox

superradcompany/microsandbox · ★ 8.3k111/7d · Rust

Microsandbox runs untrusted code in local microVMs controlled through a CLI or language SDK. It gives agent builders, plugin hosts, CI tools, and code runners stronger process separation than an ordinary container without requiring a permanent sandbox server.

SetupDocsCommunityMaturity

Verdict Our Microsandbox build failed after 347 seconds and its tests failed after 110 seconds because the linker could not find `cap-ng`. The product shape is attractive for local agent code execution, especially with SDKs across 5 languages and no daemon to operate.

Full review →
483

Dopamine

opa334/Dopamine · ★ 6.7k61/7d · C

Dopamine is a rootless, semi-untethered jailbreak for specific iPhones and iPads running supported iOS versions from 15.0 upward. It opens the device to unofficial packages and system modifications through Sileo or Zebra, while limiting writes to selected filesystem areas rather than replacing the operating system root.

SetupDocsCommunityMaturity

Verdict Dopamine is the leading option to try when its exact hardware and iOS matrix matches your device and you want a modern rootless jailbreak. Its active releases and open code inspire more confidence than an opaque download, but 3.0's fast patch cycle and serious current issue reports make a primary phone a poor first test.

Full review →
484

anti-slop

dmmulroy/anti-slop · ★ 4.5k331/7d · TypeScript

anti-slop is a set of opinionated Oxlint rules for TypeScript and JavaScript code that throws away type evidence or hides unsafe assumptions. It is distributed as source to vendor into a repository, with a Claude Code compatible agent skill that can copy and configure it.

SetupDocsCommunityMaturity

Verdict Our anti-slop run installed 11 packages in 25 seconds and passed its tests in 10 seconds, but adoption means owning 15 generic rules as vendored source. Use it when your team agrees with those exact restrictions and will tune false positives instead of accepting them as doctrine.

Full review →
485

ios-location-spoofer

mekos2772/ios-location-spoofer · ★ 4.1k214/7d · JavaScript

iOS Location Spoofer is primarily documented in Chinese; an English README exists, but it lags the Chinese file in several technical details. It rewrites Apple's network-location responses through a proxy app so an iPhone reports chosen coordinates without a jailbreak, computer, or developer account.

SetupDocsCommunityMaturity

Verdict Our worker install took 70 seconds, added 43 packages, and used 217 MB, while the repository supplied no build or test target. Use iOS Location Spoofer for authorized QA on a spare phone when you already own a compatible proxy app and accept that Apple can break the interception path.

Full review →
486

ucp

Universal-Commerce-Protocol/ucp · ★ 3.4k13/7d · Python

Universal Commerce Protocol is an open specification for platforms, merchants, payment services, credential providers, and AI agents to exchange shopping and transaction data. It defines discoverable capabilities for catalog, cart, checkout, identity, orders, and payments across REST, MCP, or A2A transports.

SetupDocsCommunityMaturity

Verdict Our UCP checkout installed 133 documentation packages in 5 seconds with 0 known npm vulnerabilities, but it exposed no build or test target and did not run a commerce transaction. UCP is worth tracking for platforms that must connect many merchants and AI agents without a custom API for each pair.

Full review →
487

cuda-oxide

NVlabs/cuda-oxide · ★ 3.3k · Rust

cuda-oxide is a compiler and runtime for writing NVIDIA GPU kernels in Rust, in the same source files as the host program. It turns those kernels into CUDA PTX and gives Rust code typed tools for memory, launches, synchronization, and asynchronous GPU work.

SetupDocsCommunityMaturity

Verdict Our cuda-oxide install added 173 packages in 91 seconds, but both build and test stopped because the sandbox had no CUDA 13.0+ toolkit. Try it if your team already owns the required NVIDIA stack and specifically wants Rust kernels with CUDA semantics.

Full review →
488

x64dbg-mcp-server

duty1g/x64dbg-mcp-server · ★ 2.0k55/7d · Zig

x64dbg-MCP Server is a native Zig plugin that lets an MCP-compatible assistant control the x64dbg Windows debugger over HTTP. Its tools can load programs, run and step them, inspect registers and memory, manage breakpoints, analyze PE files, patch a process, and save debugger data.

SetupDocsCommunityMaturity

Verdict Version 1.1 gives an MCP client 84 debugger tools over unencrypted HTTP, so x64dbg-MCP Server should be treated like a remote code-and-memory control plane, not a convenience plugin. It is a sharp tool for an isolated Windows research lab, with native deployment and unusually broad debugger coverage.

Full review →
489

gpuix

remorses/gpuix · ★ 1.9k166/7d · Rust

GPUIX lets React and TypeScript developers build desktop interfaces on Zed's native GPUI renderer instead of putting the app inside a browser window. It translates React updates into native GPU-drawn elements for macOS, Windows, and Linux, with a separate browser path through WebAssembly.

SetupDocsCommunityMaturity

Verdict Our GPUIX run installed 333 packages and used 185 MB, then the native build failed at the linker after 621 seconds, so source adoption needs a platform trial before commitment. The published quickstart is appealing for an experienced React team building a single-window tool on macOS or Windows.

Full review →
490

writer-computer

joelbqz/writer-computer · ★ 1.5k168/7d · TypeScript

Writer is a local-first macOS Markdown editor that opens ordinary folders and keeps documents on disk. It combines CodeMirror editing with rendered tables, Mermaid diagrams, KaTeX math, Obsidian-style links and images, multiple windows, Git ignore rules, and a small command-line opener.

SetupDocsCommunityMaturity

Verdict Our Writer checkout installed 683 packages in 10 seconds and occupied 522 MB, yet its root provided no build or test target that the lab could run. The v0.5.0 macOS app is appealing for people who want a focused folder-based editor with good table, math, diagram, and Obsidian-link handling.

Full review →
491

mxc

microsoft/mxc · ★ 1.3k27/7d · Rust

Microsoft eXecution Container, or MXC, gives TypeScript applications one policy format for running untrusted code through several Windows, Linux, and macOS containment backends. It targets model output, plugins, and tools, but Microsoft labels the repository an early preview and explicitly says no current MXC profile should be treated as a security boundary.

SetupDocsCommunityMaturity

Verdict Our MXC run installed 222 packages, then both build and test commands stopped because `windows-future` could not find `windows_threading::submit`, so the current source checkout is a research target rather than a drop-in containment layer. The unified schema and state-aware SDK are worth testing for cross-platform agent tooling.

Full review →
492

ASC

MG1937/ASC · ★ 1.2k · Python

Droid ASC is a Python tool for opening Android APKs, extracting one class at a time, decoding the manifest, and searching references across DEX files. It tries to spare mobile researchers from waiting for a conventional decompiler to index the whole application before the first useful search.

SetupDocsCommunityMaturity

Verdict Our ASC run installed 81 packages in 46 seconds and built in 8 seconds, but 1 of 24 tests failed on the GUI reuse path, so it is easier to trial than to trust as a settled workbench. Use it for quick, targeted APK inspection when you can compare results with another tool.

Full review →
493

Wake

iAmCorey/Wake · ★ 899161/7d · Rust

Wake is a local desktop app for browsing, searching, and reopening coding-agent conversations scattered across your machine. It combines histories from Claude Code, Codex CLI, Copilot CLI, Cursor, Gemini CLI, and several other tools without sending those records to a server.

SetupDocsCommunityMaturity

Verdict Our build failed after 896 seconds because the linker could not find `-lxkbcommon-x11`. Wake is still unusually focused and useful for developers whose agent history has become a local filing problem, especially on macOS where the project is furthest along.

Full review →
494

asyar

Xoshbin/asyar · ★ 79357/7d · Rust

Asyar is an open-source desktop launcher for macOS, Windows, and Linux, built with Rust, Tauri, and Svelte. It combines app and file search with clipboard history, scripts, extensions, window controls, AI agents, and MCP servers, so keyboard-heavy users can replace several small utilities with one launcher.

SetupDocsCommunityMaturity

Verdict Our Asyar build stopped after 9 seconds because Bun was missing, despite the 21-second dependency install succeeding, so source contributors should budget for an extra runtime prerequisite. The released app is worth trying if you want one local launcher to cover search, scripts, AI, and MCP across three desktop operating systems.

Full review →
495

pr-af

Agent-Field/pr-af · ★ 63114/7d · Go

PR-AF is a self-hosted AI pull-request reviewer that splits a change into review dimensions, asks several agents to inspect it, verifies their evidence, and posts findings back to GitHub. It is built on AgentField for teams that want deeper automated review than a diff summary and are willing to supply model and GitHub credentials.

SetupDocsCommunityMaturity

Verdict Our PR-AF run built in 4 seconds, but 22 of 107 tests failed because the test environment lacked async support, so the checkout is promising rather than release-gate ready. Trial it if deep, evidence-backed reviews justify AgentField operations and model spend.

Full review →
496

RyukGram

faroukbmiled/RyukGram · ★ 556 · JavaScript

RyukGram is a proprietary iOS tweak that changes Instagram's interface and behavior, adding downloads, feed controls, local archives, privacy switches, and extensive appearance settings. It solves annoyances that the stock app does not let users configure, but the core tweak is currently closed source despite being distributed through a public GitHub repository.

SetupDocsCommunityMaturity

Verdict Our 3-CPU, 8 GB sandbox could not run commit 953ee0d because RyukGram has no supported build ecosystem or Dockerfile. As an open-source recommendation, it fails the basic test because the core tweak is closed and its license bars modification and redistribution.

Full review →
497

real-company-interview-ai-coding-projects

CHENG-LIANG1/real-company-interview-ai-coding-projects · ★ 33512/7d

Real Company Interview AI Coding Projects is a Chinese-language study pack; its README and 5 supporting documents have no English translation. It rewrites 3 anonymized take-home assignments and adds a reusable method for scoping, building, checking, and presenting AI-assisted coding work.

SetupDocsCommunityMaturity

Verdict Our sandbox did not run commit 883fd50 because it has no supported ecosystem or Dockerfile. Use it as a Chinese-language requirements workbook, especially for its acceptance criteria and scope-cutting advice, rather than as proof that the proposed systems work.

Full review →
498

cryptolayer

igmunv/cryptolayer · ★ 32410/7d · Python

CryptoLayer is a Russian-first Python library that encrypts messages before passing them through an existing messenger, file system, or other transport. Full English README and architecture documentation exist. It supplies encryption, signatures, acknowledgments, and a module interface, leaving the application developer to provide the user interface and the actual transport.

SetupDocsCommunityMaturity

Verdict Our run passed all 6 tests and pip-audit found 0 known dependency vulnerabilities, but the project’s own open security warning says not to trust it with real confidential messages yet. CryptoLayer is worth reading and testing if you want a compact, transport-independent protocol experiment.

Full review →
499

How-To-Secure-A-Linux-Server

imthenachoman/How-To-Secure-A-Linux-Server · ★ 31k55/7d

How To Secure A Linux Server is a long, command-heavy guide to hardening a mostly Debian-style home server. It covers threat modeling, SSH, account privileges, updates, firewalls, intrusion detection, auditing, and a separate danger zone for changes that can make a machine unusable.

SetupDocsCommunityMaturity

Verdict This 171,177-character guide contains sound recovery habits and a panic-password example that deletes `/home`, so it is reference material for experienced home-server operators rather than an execution plan. Use its table of contents to find topics, then confirm each setting against current OpenSSH, Docker, kernel, and distribution documentation.

Full review →
500

genlayer-project-boilerplate

genlayerlabs/genlayer-project-boilerplate · ★ 18k220/7d · TypeScript

GenLayer Project Boilerplate is a worked starter for building an application whose contract can read the web and ask a language model to interpret what it finds. Its football betting example connects a Python contract, local tests, deployment scripts, and a TypeScript web interface, saving new GenLayer developers from assembling those pieces from scratch.

SetupDocsCommunityMaturity

Verdict Our GenLayer boilerplate install pulled 2,324 packages and 47 known vulnerabilities, including 1 critical, while the build passed and no root test target existed. Use it to learn GenLayer's contract, deployment, and frontend path, then replace the football logic and repair the dependency baseline.

Full review →
501

hetty

dstotijn/hetty · ★ 12k21/7d · Go

Hetty is a local HTTP inspection toolkit for security researchers. It puts a machine-in-the-middle proxy, request log, manual request sender, interception controls, scope rules, and a web interface into one Go application.

SetupDocsCommunityMaturity

Verdict Our Hetty run built in 6 seconds, but only 8 of 13 test targets passed, so this commit belongs on an isolated research machine rather than a shared sensitive proxy. It is appealing for a solo tester who wants readable Go code and the core capture, intercept, and replay loop.

Full review →
502

cordis

cordiverse/cordis · ★ 8.5k269/7d · TypeScript

Cordis is a TypeScript plugin framework for services that appear, disappear, reload, and depend on one another while a process is running. Its main README is English, and the linked documentation defaults to Chinese but includes a full English section covering contexts, services, events, fibers, and plugin development.

SetupDocsCommunityMaturity

Verdict Our Cordis checkout installed 432 packages with 0 known vulnerabilities, but both build and test stopped within 10 seconds because the npm-created workspace state did not satisfy Yarn 4.14.1. The framework is interesting for agent hosts and other long-lived systems that need dependency-aware plugins and reversible side effects.

Full review →
503

defending-code-reference-harness

anthropics/defending-code-reference-harness · ★ 7.5k66/7d · Python

Anthropic's Defending Code Reference Harness is an example security workflow built around Claude Code skills and autonomous agents. It covers threat modeling, static review, crash verification, deduplication, reporting, and patch checking, with a reference pipeline aimed at C and C++ memory vulnerabilities.

SetupDocsCommunityMaturity

Verdict Our harness build passed in 8 seconds and 411 of 426 tests passed, while 14 failures and 1 setup error all stopped on a missing `docker` executable. Use this repository as design material for a staffed security program that can own sandboxing, target configuration, model access, and human review.

Full review →
504

vite-plus

voidzero-dev/vite-plus · ★ 5.8k34/7d · Rust

Vite+ is a single `vp` command for managing Node.js, dependencies, development, checks, tests, application builds, library packaging, and cached monorepo tasks. It bundles Vite, Vitest, Oxlint, Oxfmt, Rolldown, tsdown, and Vite Task so a web project can keep one typed configuration and one toolchain entry point.

SetupDocsCommunityMaturity

Verdict Our Vite+ source install stopped after 16 seconds on a missing Chokidar patch, so commit 3380eb8 is not a dependable contributor baseline. Trial the released `vp` binary on a disposable branch if consolidating the Vite stack would remove real maintenance from your team.

Full review →
505

exploitarium

bikini/exploitarium · ★ 5.1k67/7d · Python

Exploitarium is a consolidated archive of public vulnerability research and proof-of-concept code covering many unrelated software projects. It helps security researchers study reported flaws, reproduce findings in controlled environments, and inspect the evidence behind a disclosure, but it is not a single ready-to-run application.

SetupDocsCommunityMaturity

Verdict Our install finished in 17 seconds and the build in 10 seconds. Exploitarium is useful as a research shelf for experienced defenders, especially because it preserves many PoCs with explicit provenance, but it is not a dependable package or turnkey toolkit.

Full review →
506

skills

humanlayer/skills · ★ 4.1k563/7d · TypeScript

HumanLayer Skills is a small collection of Claude Code instructions for editing CLAUDE.md files, tightening React props, designing recurring coding-agent jobs, and explaining code visually. It gives an agent detailed working methods rather than shipping a library or service that your application calls.

SetupDocsCommunityMaturity

Verdict Our sandbox did not run humanlayer/skills because it exposed no supported ecosystem and had no Dockerfile, so there is no executable result behind this review. Use the five skills as editable methods for Claude Code, especially the React prop audit and bounded agent-loop templates.

Full review →
507

hop

golbin/hop · ★ 1.8k15/7d · TypeScript

HOP is a Korean-language desktop editor for opening HWP and HWPX documents on macOS, Windows, and Linux. Its README and development guide are in Korean, and neither links to a full English documentation set. The app wraps the rhwp document engine with file handling, PDF export, printing, and desktop integration.

SetupDocsCommunityMaturity

Verdict Our Hop run installed 54 packages in 6 seconds and passed the JavaScript tests, but the full command failed because Cargo was absent. HOP is worth trying when HWP access on macOS or Linux is the immediate problem and your workflow can stay with HWP for saving.

Full review →
508

typephp

swoole/typephp · ★ 1.4k69/7d · PHP

TypePHP is an ahead-of-time compiler that turns a supported subset of PHP into native executables, PHP extensions, shared libraries, or WASI components. It aims to give typed, computation-heavy PHP code native compilation without forcing developers to rewrite an entire project in C++.

SetupDocsCommunityMaturity

Verdict We could not produce an install, build, or test result because our sandbox had no supported PHP runner and the repo had no Dockerfile. TypePHP is unusually ambitious and technically specific, but version 0.6.6 still reads like an expert-led evaluation target rather than a safe default for production PHP.

Full review →
509

herdrm

missuo/herdrm · ★ 68924/7d · Swift

herdrm is a native macOS console for watching and entering coding-agent terminals managed by the separate herdr daemon. It gathers Claude Code, Codex, Gemini, Grok, OpenCode, and other command-line agents from a local Mac and remote SSH machines into one sidebar, search palette, and terminal window.

SetupDocsCommunityMaturity

Verdict herdrm requires macOS 14 or newer, a herdr daemon on each machine, and agent bypass permissions are enabled by default, so adoption should begin on non-sensitive development hosts. It is a compelling console for people already juggling several herdr-managed agents, especially across SSH boxes.

Full review →
510

Backend-from-first-Principle

DsThakurRawat/Backend-from-first-Principle · ★ 49036/7d · JavaScript

Backend from First Principles is an English-language course made of 24 browser-readable chapters on HTTP, APIs, data stores, reliability, security, deployment, and distributed communication. It pairs the explanations with small examples in Go, Python, SQL, shell, configuration formats, and some JavaScript rather than building one application from start to finish.

SetupDocsCommunityMaturity

Verdict Our lab produced no install, build, or test result because commit 6e2d007 had an unsupported JavaScript layout and no Dockerfile. Use Backend from First Principles as a 24-chapter reading map, especially if Go and Python examples suit you, while checking primary documentation before applying security or platform guidance.

Full review →
511

noty

aimen08/noty · ★ 440 · Swift

Noty is a Swift macOS app with English documentation that keeps sticky notes in a narrow deck along one screen edge. It gives you quick capture, inline tasks and images, local encrypted storage, optional iCloud Drive mirroring, and a URL scheme for automation.

SetupDocsCommunityMaturity

Verdict We did not run Noty at commit f51dd30 in our 3-CPU Debian sandbox: its Swift macOS ecosystem was unsupported and the repository had no Dockerfile. Try v1.8.0 if you use macOS 15 and the 56-point edge deck solves a real capture problem for you.

Full review →
512

dsh-explore

antinomie-lab/dsh-explore · ★ 178 · Vue

Deepseek Harness Explore is primarily a Chinese-language reading and research project, with an English README available. It explains how Cordis and Deepseek Harness organize agent-harness components through plugin trees, lifecycle rules, and declarative configuration, while providing a Vue reading site for the material.

SetupDocsCommunityMaturity

Verdict Use Deepseek Harness Explore as a focused technical reading resource, not as the foundation of a production system. Its source-pinned citations and concrete Cordis configuration topics make it unusually promising for architecture research, but one listed article, no releases, and an unknown license make it too early to treat as a dependable product.

Full review →
513

forward-implementation-first

Vuk97/forward-implementation-first · ★ 169 · Shell

Forward implementation first is a Markdown skill for Claude Code, Codex, and other coding agents. It tells an agent to keep building and checking outputs when stale hashes, receipts, locks, or progress records would otherwise send the work backward.

SetupDocsCommunityMaturity

Verdict Our sandbox could not execute commit 2d4dd7e because this Shell repository has no supported build ecosystem or Dockerfile. Install it only if a long pipeline has already lost time to stale receipts, hashes, or status records, and review the hard constraints against your own safety controls first.

Full review →
514

paper

cordiverse/paper · ★ 3.0k9/7d

cordiverse/paper is an 88-page English preprint about making software components safe to add, remove, and reconfigure while a program keeps running. It proposes runtime bookkeeping for undoing a component's changes and for reacting when its dependencies appear or disappear, then relates the theory to the separate Cordis TypeScript framework.

SetupDocsCommunityMaturity

Verdict Read this paper if you are designing a serious plugin host or self-modifying agent runtime, because it gives unusually precise names and rules to problems that are often handled with restart hooks. Do not mistake the proof for a ready-made platform or measured production result.

Full review →
515

darwin-vm

jprx/darwin-vm · ★ 1.1k46/7d · Python

darwin-vm is a QEMU-based environment for booting recent iOS and macOS kernels into a bare root shell. It gives kernel researchers a place to inspect Darwin internals, patch a recovery filesystem, run command-line programs, and attach a debugger without using a physical iPhone or Mac as the target. It does not reproduce the screen or the rest of the consumer device experience.

SetupDocsCommunityMaturity

Verdict Our darwin-vm checkout installed 60 Rust packages in 9 seconds, but both the 10-second build and 4-second test run stopped because `MESON_BUILD_ROOT` was absent. Use it if recent Apple kernel research is the job and you can follow the full QEMU and macOS preparation path.

Full review →
516

babua-dsa-patterns-course

team-codebug/babua-dsa-patterns-course · ★ 95814/7d · JavaScript

Babua DSA Patterns Course is a Hindi-explained coding-interview course, with an English README and onboarding sheet. It organizes LeetCode practice around recurring problem-solving patterns, pairing YouTube lessons with day folders, visual notes, and tracking sheets.

SetupDocsCommunityMaturity

Verdict Use Babua DSA Patterns Course if Hindi instruction and a fixed daily sequence help you keep practicing. Its videos, visual notes, and revision sheet make a workable guided course, but the repository is an uneven companion archive rather than a complete learning product on its own.

Full review →
517

gotohp

xob0t/gotohp · ★ 592 · Go

gotohp is an unofficial desktop and command-line uploader for Google Photos. It handles files or whole directories, can skip media already in an account, and gives users live progress and configurable upload threads without relying on the Google Photos website.

SetupDocsCommunityMaturity

Verdict Our gotohp build took 6 seconds, but all 7 `go test` targets failed after 45 seconds because two Linux libraries were unavailable. Try it for a one-way Google Photos upload job only if you accept the unofficial credential flow and can verify the result in your account.

Full review →
518

refactoring-ui-skill

s0xDk/refactoring-ui-skill · ★ 55012/7d · CSS

Refactoring UI Skill is a Claude Code instruction pack that turns lessons from Adam Wathan and Steve Schoger's Refactoring UI into fixed rules for spacing, typography, color, shadows, and visual hierarchy. It helps Claude make fewer arbitrary styling choices when building or repairing an interface.

SetupDocsCommunityMaturity

Verdict Our sandbox did not run commit 4887214 because this CSS-classified repository has no supported build ecosystem or Dockerfile, so adoption rests on reading the instructions and reviewing Claude's output. Use it if your Claude Code UI work needs firm scales and a practical token baseline.

Full review →
519

cyclomatic-complexity-skill

saurabhkumar8112/cyclomatic-complexity-skill · ★ 39011/7d

Cyclomatic Complexity is a Claude Code skill that tells the agent to measure branching in functions, refactor the worst cases, and report the before-and-after counts. It packages a sensible code-review routine as written instructions rather than shipping a new analyzer.

SetupDocsCommunityMaturity

Verdict Our sandbox could not run commit 567886f because the repository has no supported ecosystem or Dockerfile, so there is no execution evidence behind this skill yet. Use it as a short Claude Code checklist if your project already has a complexity analyzer and tests.

Full review →
520

typescript-go

microsoft/typescript-go · ★ 26k · Go

TypeScript Go is the staging repository Microsoft used to build the native Go port of the TypeScript compiler and language service. That work has moved into microsoft/TypeScript, and this repository now tells developers to continue there before it is archived in September 2026.

SetupDocsCommunityMaturity

Verdict Our source build stopped after 8 seconds because `go` was missing, and the repository itself is due for archival in September 2026, so new users should go straight to microsoft/TypeScript. Keep this checkout only for historical investigation or work tied to the old `@typescript/native-preview` line.

Full review →
521

my-girlfriend-jingtian-latex

HEJustinSun/my-girlfriend-jingtian-latex · ★ 4.2k · TeX

My Girlfriend Jingtian is a Chinese-language XeLaTeX project for one specific short book, and it has no English documentation. It gives readers the finished PDF and gives TeX users the source, cover, and bundled Chinese fonts needed to study or modify its 5 x 8 inch layout.

SetupDocsCommunityMaturity

Verdict This repository gives you one 5 x 8 inch Chinese book and its XeLaTeX source, but it provides no English documentation, release, or declared project license. Read the PDF or borrow a layout technique if the terms fit your use; do not mistake it for a supported publishing framework.

Full review →
522

Open_Duck_Mini

apirrone/Open_Duck_Mini · ★ 4.1k74/7d · Python

Open Duck Mini is an open hardware project for building a 42-centimeter biped inspired by Disney's BDX droid. The repository collects printable parts, CAD, wiring, assembly notes, simulation files, and pretrained walking policies, while the onboard runtime and current training environment live in separate repositories.

SetupDocsCommunityMaturity

Verdict Our Open Duck Mini checkout installed 36 packages and built in 6 seconds, but it offered no test target and did not validate a physical robot. Build it if the mechanical work and sim-to-real tuning are the point of the project.

Full review →
523

everything-claude-code

WorldFlowAI/everything-claude-code · ★ 3.0k259/7d · JavaScript

Everything Claude Code is a bundle of agents, skills, commands, rules, hooks, and MCP configuration examples for customizing Claude Code. This WorldFlowAI repository is a January 2026 snapshot whose README and plugin metadata still identify the original project, now hosted at affaan-m/ECC, so it solves the setup problem with a significant source and maintenance caveat.

SetupDocsCommunityMaturity

Verdict Do not install the WorldFlowAI snapshot as your default Claude Code toolkit. It contains useful patterns and readable examples, but its stale default branch, upstream-facing identity, and unresolved discovery defects make the maintained affaan-m/ECC repository the obvious first stop.

Full review →
524

skitter-creek-bath-salts

xoreaxeaxeax/skitter-creek-bath-salts · ★ 2.0k45/7d · C

skitter-creek-bath-salts is a security-research project that changes how certain older AMD processors translate physical memory addresses into DRAM locations. By disturbing that final mapping layer, it demonstrates a path to memory regions that normal CPU and firmware protections are supposed to hide, including PSP, System Management Mode, C6 storage, and microcode-related areas.

SetupDocsCommunityMaturity

Verdict Use skitter-creek-bath-salts if you are doing serious platform-security research on disposable AMD Family 16h hardware and its unusually clear explanation is as valuable to you as the code. Everyone else should read it, learn from its model of the memory stack, and stop before running it.

Full review →
525

Antigravity

confeden/Antigravity · ★ 53962/7d · Rust

Antigravity is a Russian-language Windows unlocker for using Google's Antigravity coding tools from regions that Google rejects. The project does not provide English documentation. It patches one field in Google's installed binary and changes selective Windows DNS settings so affected Google endpoints take a different route.

SetupDocsCommunityMaturity

Verdict Our Antigravity build passed in 35 seconds, but 4 of 224 tests failed in the 13-second suite. Its selective routing may suit a Russian-speaking Windows user blocked by Google's region check, provided that modifying a vendor binary and Windows DNS policy is acceptable.

Full review →
526

dsh-gitbash-preset

liceses/dsh-gitbash-preset · ★ 131 · JavaScript

dsh-gitbash-preset is a DeepSeek Harness plugin whose main README is written in Simplified Chinese; no separate English guide is linked, though package metadata has an English description. It adds a Windows version of DSH's minimal agent preset, routing each bash tool call through Git for Windows when DSH's normal persistent shell cannot run on Win32.

SetupDocsCommunityMaturity

Verdict Our run installed 0 dependencies in 5 seconds, yet 3 of the 11 counted tests failed, so this tiny plugin still needs a local compatibility check before use. It is a narrow answer for Chinese-reading DeepSeek Harness users who already have Git for Windows and can approve full-access commands.

Full review →
527

cve-2026-41940-PoC

lanicer/cve-2026-41940-PoC · ★ 532 · Python

This English-language Python proof of concept automates exploitation of CVE-2026-41940 against cPanel and WHM. It goes beyond checking exposure: the documented commands can obtain privileged access, read server data, execute operating-system commands, change the root password, and create accounts.

SetupDocsCommunityMaturity

Verdict Our run installed 35 packages in 11 seconds and built in 3 seconds, but there was no test target, so this PoC is easy to prepare and unproven as a diagnostic. Use it only for an authorized, isolated reproduction where exploit behavior is the point.

Full review →
528

Dalaran

Flaminis/Dalaran · ★ 776 · Rust

Dalaran is a self-hosted viewer, recorder, and query stack for time-aligned robot data such as camera frames, LiDAR, IMU readings, and transforms. It is a hard fork of Rerun that adds Python, Rust, and C++ logging APIs while keeping compatibility with the upstream recording container.

SetupDocsCommunityMaturity

Verdict Our Dalaran run built in 9 seconds, but pytest produced 29 failures and 96 collection or setup errors, so this fork is a research trial rather than a safe default for a robotics team. Its local-first licensing and upstream recording compatibility are attractive.

Full review →
529

HardBreacher

MSNightmare/HardBreacher · ★ 276 · C

HardBreacher is a Windows privilege-escalation proof of concept aimed at Kaspersky Endpoint Security. It demonstrates a claimed flaw that can give the current user control over a file placed in Windows System32; it is not a scanner, patch, or defensive monitoring tool.

SetupDocsCommunityMaturity

Verdict Our lab result for HardBreacher is `NOT RUN`: the harness does not support its C ecosystem, and the repository has no Dockerfile. Use the code only as unverified research material in an authorized, disposable Windows lab.

Full review →
530

IKONA-Security

KorekKayu/IKONA-Security · ★ 110 · Python

IKONA Security is an Indonesian-first collection of offensive-security notes and tools; the same README includes English, Chinese, and Russian sections. Its folders mix web-audit playbooks, agent skill files, payload references, wordlists, and Python probes for authorized security work.

SetupDocsCommunityMaturity

Verdict Our 2026-09-16 lab did not run IKONA Security because Python was outside this batch's supported ecosystems and the repository had no Dockerfile, so there is no verified install, build, or test result. Treat it as raw research material, not as a ready security suite: select one file, inspect it, pin its dependencies, and use an authorized lab target.

Full review →
531

tokentab

damejan80/tokentab · ★ 1.2k496/7d · Python

Tokentab presents itself as a local cost report for Claude Code, Codex, and Gemini CLI session logs. The checked source does contain log parsers and an offline dashboard, but its command entry point also calls an unrelated Windows-only downloader that can execute Python fetched from a hard-coded IP.

SetupDocsCommunityMaturity

Verdict Our Tokentab install finished in 22 seconds and built in 3 seconds, but commit 80358bc contains a Windows-only path that downloads Python from a hard-coded IP and executes it in memory. Do not install or run this commit on a real workstation.

Full review →
532

CVE-2026-9830

opaxial/CVE-2026-9830 · ★ 462 · Python

This repository presents itself as a Python checker for CVE-2026-9830, an unauthenticated BookingPress Pro API flaw affecting versions before 5.7.3. The checked-in script also hides code that downloads and executes text from a bare IP address, so it is unsafe to run.

SetupDocsCommunityMaturity

Verdict Static inspection of our 373-line checkout found an obfuscated statement that fetches and executes remote text from `5.255.114.98:8090`, so nobody should run this repository as supplied. The underlying CVE is real, and NVD says BookingPress Pro versions before 5.7.3 are affected, but this proof of concept creates a separate compromise path on the assessor's machine.

Full review →