Today's Open Source Growth Pulse

2026-10-30 · 20 projects · 24h star increment tracking

This issue selects 20 open-source projects worth attention based on the last 24 hours of star increments. Specific capabilities are subject to the project repositories.

Today in 30 seconds

Rank	Project	24h Increment	Language
🥇 1	morluto/rea	+5,392	TypeScript
🥈 2	pbakaus/impeccable	+2,591	Rust
🥉 3	Ebony-Vinyl/dsh-our-free-model	+2,092	JavaScript
4	HKUDS/DeepTutor	+2,032	Python
5	togg53192-cmd/jailbreaks	+1,897	Unknown language
6	Gathin23/cleanyourtv	+1,815	Python
7	Nutlope/hallmark	+1,513	CSS
8	debpalash/VoiceStudio	+1,408	Python
9	hmjz100/LinkSwift	+1,367	JavaScript
10	cloudflare/security-audit-skill	+1,290	JavaScript
11	google/elephant-goldfish-model	+1,266	Python
12	anhao/ai-pose-gallery-zh	+1,254	HTML
13	obra/superpowers	+1,187	Shell
14	stablyai/orca	+1,156	TypeScript
15	interplains/Zero-Breach-Toolkit	+1,138	Python
16	KnifeLemon/Filee	+1,134	C#
17	elayadesign/ai-design-skills	+1,118	Unknown language
18	DietrichGebert/ponytail	+1,064	JavaScript
19	ifixai-ai/iFixAi	+1,049	Python
20	unclebob/ub-mode	+1,040	Unknown language

Biggest increment today: +5,392 (morluto/rea) · Top 3 growth combined: +10,075

Project deep dive

🥇 01｜rea ⭐ 407,645 (24h +5,392)

REA is a reverse engineering tool built on TypeScript that provides a unified MCP (multi-function plugin) framework for reverse analysis of binaries, applications, and runtime behavior. Through pluggable agent modules, it helps users quickly locate feature implementations and understand internal mechanics, and can be used directly on multiple platforms (such as Hopper). It serves security researchers, software reverse engineers, and developers, tackling reverse engineering problems from high-level functionality down to low-level binaries.

Project page: github.com/morluto/rea

🥈 02｜impeccable ⭐ 95,236 (24h +2,591)

Impeccable is a frontend design guidance system for AI coding assistants. Through 23 commands, 59 deterministic detection rules, and real-time browser iteration, it helps AI-generated frontend interfaces break free from homogenized design templates (such as the Inter font, purple-blue gradients, and nested cards), providing developers and AI coding tools (Cursor, Claude Code, GitHub Copilot, etc.) with a structured design language and quality review capability.

Project page: github.com/pbakaus/impeccable

🥉 03｜dsh-our-free-model ⭐ 50,855 (24h +2,092)

dsh-our-free-model is a plugin for the dsh platform that provides free, unlimited access to frontier large models such as DeepSeek V4.1 Flash and Kimi K3, with no login, registration, or API key required. It mounts directly with zero dependencies and no build steps, and is suitable for users who need to quickly use high-performance LLMs in local or private environments.

Project page: github.com/Ebony-Vinyl/dsh-our-free-model

04｜DeepTutor ⭐ 49,959 (24h +2,032)

DeepTutor is a lifelong personalized tutoring platform built on Python and Next.js. By integrating multiple large language models, knowledge base retrieval (RAG), and intelligent tutoring agents, it provides learners with personalized tutoring, knowledge management, guided learning paths, and multimodal interaction capabilities. It targets educators and self-directed learners who need adaptive learning support.

Project page: github.com/HKUDS/DeepTutor

05｜jailbreaks ⭐ 8,928 (24h +1,897)

This project collects 31 system prompts, role definitions, workspace policies, and agent operational norms targeting 25+ models, covering free invisible models, open-weight models, and closed frontier models. It aims to help researchers and developers explore how prompt hierarchy, role frameworks, and policy coverage affect model behavior, promoting AI system resistance to prompt injection and instruction manipulation.

Project page: github.com/togg53192-cmd/jailbreaks

06｜cleanyourtv ⭐ 45,434 (24h +1,815)

tv-vibehack is an ADB-based tool that helps users remove ads and preinstalled bloatware from Android TV/Google TV without root or physical connection, and install a clean launcher that looks similar to Apple TV. It is suitable for ordinary home users and tech enthusiasts.

Project page: github.com/Gathin23/cleanyourtv

07｜hallmark ⭐ 42,056 (24h +1,513)

Hallmark is a design skill pack for AI code editors (Claude Code, Cursor, Codex). Through 57 anti-pattern detections and 21 theme systems, it generates UI code with no trace of AI. It automatically selects macro structure, color scheme, and typography based on different needs, ensuring each generated page has a unique design fingerprint rather than templated output.

Project page: github.com/Nutlope/hallmark

08｜VoiceStudio ⭐ 94,244 (24h +1,408)

VoiceStudio is an open-source, fully local voice cloning and design tool that supports voice cloning in 646 languages, voice design, video dubbing, dictation, transcription, and audiobook production. It includes a desktop Electron app and a local API, and is suitable for creators, developers, and enterprises.

Project page: github.com/debpalash/VoiceStudio

09｜LinkSwift ⭐ 36,222 (24h +1,367)

LinkSwift is a JavaScript-based cloud drive direct link tool, modified from the “网盘直链下载助手” (cloud drive direct link download assistant). It supports eight major cloud drives including Baidu, Aliyun, Mobile, Tianyi, Xunlei, Quark, UC, and 123, helping users quickly obtain direct download links for files. It suits individual users who need batch or manual downloads from cloud drives.

Project page: github.com/hmjz100/LinkSwift

10｜security-audit-skill ⭐ 52,805 (24h +1,290)

security-audit is a code agent skill used inside Cloudflare that provides a multi-stage, machine-readable security audit process. Through six independent stages (reconnaissance, coverage-driven hunting, candidate verification, structured output, independent record verification, target-neutral reporting), it generates verifiable findings files, and is suitable for development and security teams that need systematic, traceable vulnerability discovery.

Project page: github.com/cloudflare/security-audit-skill

11｜elephant-goldfish-model ⭐ 5,884 (24h +1,266)

Elephant-Goldfish Model (EGM) is an AI plugin open-sourced by Google. It helps developers collect project context and draft design specifications through the “Elephant” role before coding, after which the “Goldfish” role reviews the draft with zero context, revealing potential assumptions and blind spots. It suits feature development that needs early design decisions.

Project page: github.com/google/elephant-goldfish-model

12｜ai-pose-gallery-zh ⭐ 3,107 (24h +1,254)

AI Pose Gallery Chinese Edition is a zero-dependency responsive static webpage that provides 1,439 entries across 288 sets of AI human pose reference images (5 viewpoints per image), fully localized into Chinese and directly downloadable. It serves as reference material for AI painting, 3D modeling, photography, and other creative work.

Project page: github.com/anhao/ai-pose-gallery-zh

13｜superpowers ⭐ 308,496 (24h +1,187)

Superpowers is a software development methodology framework for AI coding agents. Through a composable skill system, it guides agents to work in a standardized process, from requirements clarification, design review, and implementation planning to test-driven development and code review. It enforces systematic engineering practices and supports multiple mainstream AI coding tools.

Project page: github.com/obra/superpowers

14｜orca ⭐ 110,154 (24h +1,156)

Orca is an AI agent orchestration desktop environment (ADE) for efficient developers. It supports running multiple CLI coding agents in parallel (such as Claude Code, Codex, etc.), isolates each agent task through independent Git worktrees, and tracks and manages them in a unified interface. It suits professional developers and engineering teams that want to significantly improve coding efficiency in an “agent fleet” mode.

Project page: github.com/stablyai/orca

15｜Zero-Breach-Toolkit ⭐ 3,572 (24h +1,138)

Zero Breach Toolkit is a free, open-source offline tool that uses Python to detect traces of known commercial spyware (such as Pegasus, Predator, etc.) in Windows, Mac, Android, and iPhone backups. It is read-only, never modifies the system, reports detection results, and explains the strength of evidence. It suits journalists, activists, and human rights organizations that need initial self-check security assessments.

Project page: github.com/interplains/Zero-Breach-Toolkit

16｜Filee ⭐ 17,392 (24h +1,134)

Filee is a free, open-source offline file conversion tool that supports 180+ formats, including images, PDF, Office documents, ebooks, audio, and video. Users drag a file with a key press at the cursor position, then select the target format directly from a format ring that pops up. Everything runs locally, with no extra software or network uploads.

Project page: github.com/KnifeLemon/Filee

17｜ai-design-skills ⭐ 12,090 (24h +1,118)

AI Design Skills is a library of design skills for AI coding tools (such as Claude Code, Cursor, Codex, Windsurf, etc.). Each skill lives in its own folder, and users can copy the corresponding SKILL.md rule file as needed to let AI automatically complete design tasks such as building a landing page from scratch. It suits developers and designers who need to quickly generate or upgrade web designs.

Project page: github.com/elayadesign/ai-design-skills

18｜ponytail ⭐ 180,138 (24h +1,064)

Ponytail is an open-source project that aims to make AI agents think like a lazy senior developer, reducing unnecessary code writing and improving development efficiency. The project is written in JavaScript and has over 100,000 stars, so it is worth attention.

Project page: github.com/DietrichGebert/ponytail

19｜iFixAi ⭐ 37,863 (24h +1,049)

iFixAi is an independent audit tool for AI agents that verifies whether agents correctly execute tasks according to business KPIs and organizational architecture requirements through 49 checks in 120 seconds. It provides three execution modes, a CLI wizard, explicit commands, and plugins, and supports cross-vendor evaluation mechanisms to ensure audit results are referenceable. It targets developers and enterprises that need to quickly verify agent behavioral compliance.

Project page: github.com/ifixai-ai/iFixAi

20｜ub-mode ⭐ 4,666 (24h +1,040)

ub-mode is a set of agent skills for code writing and modification that follows strict software engineering principles. Through the three-stage process of “make it work first, then make it right, practice continuously,” combined with deterministic quality gate tools such as CRAP, duplication metrics, and mutation testing, it helps developers improve code quality and maintainability. It suits individuals or teams that need strict code quality control.

Project page: github.com/unclebob/ub-mode

---

Data source: YourTools GitHub trending board, updated daily.

Original article: https://tools.wflynn.cn/aitools/github-trending

Follow the WeChat official account 「YourTools」for daily GitHub trends and open source tool digests: