GitHub is an online warehouse where people put program code and share it. Each program lives in its own box called a repository — a repo for short — and each card here is one repo.
These are repos created in the last two weeks that already have more than 30 stars, taken most-starred first. In short: brand new, and already drawing attention.
Stars count people who clicked “I like this, save it for later”; forks count people who copied the whole thing to change it their own way. In the grey line, the licence is the terms for reusing it, and “pushed” is when the code was last changed.
What turned up in the last three days — collected by machine, published by hand.
What gets picked here ends up in What I took apart →
Collected 2026-10-15 23:01 UTC · 44 kept from 418
2.4k stars322 forks
MITPythonpushed today
Location, estimated · United States
“Let an AI model use a real iPhone with a $7 ESP32: AirPlay eyes, USB HID hands, any OpenAI-compatible vision model. No Mac, no …”
Published 2026-10-11
Fetched 2026-10-13 23:01 UTC · 56e3d501b6411969
What the account wrote — New York
Torn from — the README
2.0k stars192 forks
Apache-2.0JavaScriptpushed today
Location, estimated · Not stated
“A team of long-lived bots for DeepSeek Harness, with main bots, bot-to-bot messages and group chats.”
Published 2026-10-07
Fetched 2026-10-13 23:01 UTC · f95ff77c0b1f0b50
What the account wrote — The account left it blank
Torn from — the README
2.7k stars281 forks
Apache-2.0Rustpushed today
Location, estimated · Not stated
“SolidCraft: an open-source parametric 3D CAD application built in pure Rust. Desktop, web, and automation through MCP.”
They claim — A Raspberry Pi 4 case from an empty design: a sketch dimensioned with the board's parameters, extrude, fillet, shell, the rim rounded in one pick, an M2.5 standoff patterned on the 58 × …
Published 2026-10-07
Fetched 2026-10-14 23:02 UTC · edd2f0662ef18e33
What the account wrote — The account left it blank
Torn from — the README
1.5k stars1.8k forks
AGPL-3.0Gopushed today
Location, estimated · Korea
“ARTEX 한국어판 · AI 자율 침투 테스트 프레임워크(upstream: Autumn-27/ARTEX, AGPL-3.0)”
They claim — A cheaper, smaller model (for example gpt-4o-mini), by contrast, let the report fall back to the original language.
Published 2026-10-03
Fetched 2026-10-13 23:01 UTC · ba08d341154c8561
What the account wrote — South Korea
Torn from — the README
2.0k stars414 forks
no licensePythonpushed today
Location, estimated · Not stated
“Qwen-Image 2.1 NSFW/姿势出图 skill：作者级提示词配方 + NSFW LoRA + 136 姿势 VNCCS 人偶控制；适用 Claude C…”
They claim — 底模为 Qwen-Image 2.1 int8 套件，模型文件合计约 16GB（diffusion 6.8GB / text encoder 8.7GB / VAE 0.6GB），LoRA 各自 ≤608MB。
Published 2026-10-13
Fetched 2026-10-15 23:01 UTC · 6dfee4bf4e7948d1
What the account wrote — The account left it blank
Torn from — the README
1.4k stars99 forks
MITPythonpushed today
Location, estimated · United States
“The animation engine for agents: blazingly fast math and 3D videos with Manim's Python API, on a Rust GPU core.”
They claim — Across five scenes at 1080p60 on an Apple M4 Pro, ManimGX is 90.6× faster than ManimCE, 11.5× faster than ManimGL, and 18.7× faster than Blender Workbench, from launch to finished MP4.
Published 2026-10-05
Fetched 2026-10-15 23:01 UTC · f52bdc10842e8288
What the account wrote — United States of America
Torn from — the README
Nothing on this page matches what you picked.
Papers, repos, models, blogs, Reddit and podcasts — once a day, by machine. Everything collected gets taken apart once — the README, the abstract, the model card — and what that turns up is kept for good, whether or not it becomes a card. Titles and links go to the original; nothing is reproduced here.
The grey line on each card is what the teardown found, not what the source advertises: the licence, the last commit, the model it was built on, whether the paper states its own limits. Open “What we saw” to check what was read to get it. Where it says listing only, the original could not be read — some sites refuse us, and we do not pretend otherwise.
A line in quotation marks is a sentence lifted from the source itself — their number, not a measurement of ours. We copy it and leave it unchecked; that is the point of quoting it.
Topics are assigned by matching words in the original title, quote and tags against a fixed list — no model reads the card. Some cards match nothing and carry no topic at all; pick a topic and those drop out. Source and topic narrow together.
Each card keeps the sentence we actually read at the moment we read it, plus a fingerprint of the response we received. That sentence — not the whole page — is what we can still stand behind if the original later changes or disappears.
Guessed from one free-text field the account filled in itself, and half of them leave it empty. Shown as its own value rather than hidden.