This week's top 10
-
morluto/rea
One plug-in that lets your AI agent take apart any app or program and explain how a feature works, right down to the compiled code.
Why it's hot +248,962 stars this week, 294k total, picked up by one AI source we track · TypeScript
Who should care Security researchers, app developers studying competitors, and anyone curious how software really works.
-
debpalash/VoiceStudio
A free desktop app that clones voices, dubs videos, reads out audiobooks and transcribes speech, all running on your own computer instead of a paid cloud service.
Why it's hot +14,708 stars this week, 72k total · Python
Who should care Video creators, podcasters and anyone paying for a voice cloning subscription.
-
mattpocock/skills
A small set of reusable instructions that make coding agents work like a careful senior engineer instead of guessing.
Why it's hot +11,937 stars this week, 294k total · Shell
Who should care Developers who use AI coding tools daily and want more reliable output.
-
cathrynlavery/diagram-design
Ask your agent for a diagram in plain words and get a clean, on-brand chart as a single web file, with 45 diagram types to pick from.
Why it's hot +9,024 stars this week, 57k total · HTML
Who should care Anyone who makes architecture diagrams, slides or docs and is tired of ugly auto-generated charts.
-
rehan-remade/universal-modder
A toolkit that lets an AI coding agent mod almost any PC game you own: it reads the game's code, builds the mod, makes the art and sound, and tests it.
Why it's hot about +476 stars a day since launch, 8.1k total, launched 17 days ago, picked up by one AI source we track · Python
Who should care Gamers and modders who want custom content without learning each game's internals.
-
ifixai-ai/iFixAi
A two-minute test that checks whether an AI agent actually does the job it was given, judged against your business goals, not just speed or token use.
Why it's hot +6,913 stars this week, 29k total · Python
Who should care Companies putting AI agents into real workflows and anyone who has to sign off on them.
-
onehumanbeing/esp32-iphone-agent
Plug a $7 microcontroller into an iPhone and an AI model can see the screen and tap through apps for you, with no jailbreak and no computer.
Why it's hot 2,892 stars since launch, launched 6 days ago · C
Who should care Hardware tinkerers and anyone testing what phone-using AI agents can really do.
-
heygen-com/hyperframes
Turn a web page with animations into a finished MP4 video, and let your coding agent write those pages for you.
Why it's hot +5,601 stars this week, 65k total · TypeScript
Who should care Marketers and developers who make product videos, explainers or social clips.
-
stablyai/orca
A desktop app for running several coding agents side by side, each on its own copy of your code, with a phone app to check in on them.
Why it's hot +1,102 stars today, 95k total · TypeScript
Who should care Developers juggling multiple AI coding agents at once.
-
affaan-m/ECC
An add-on pack that gives coding agents memory, security checks and a research-first way of working.
Why it's hot +1,057 stars today, 282k total · JavaScript
Who should care Developers who want to squeeze more reliable work out of their coding agent.
Moving fastest today
- morluto/rea One plug-in that lets your AI agent take apart any app or program and explain how a feature works, right down to the compiled code.
- debpalash/VoiceStudio A free desktop app that clones voices, dubs videos, reads out audiobooks and transcribes speech, all running on your own computer instead of a paid cloud service.
- xerj-org/xerj A local search engine for your code and files, so a coding agent jumps straight to the right snippet instead of reading files one by one and burning tokens.
- anthropics/claude-plugins-official Anthropic's own catalog of vetted add-ons for Claude Code, installable with one command.
- feder-cr/git-ai An open-source version control server built for hundreds of coding agents saving work at once, modeled on a design Anthropic described but never released.
How we rank
Every day we snapshot star counts for a few hundred repos from GitHub's trending lists and the newest AI projects, keep only AI projects, and score each one on star growth, how many of the AI newsletters, forums and feeds we read are talking about it, and how new it is. Then we read each README and sum it up in one plain sentence. Star farms and sketchy projects get dropped.