05

GitHub

Page 2 of 3

These are earlier cards, kept as text. Pictures and short summaries stay only while a card is on the first page; the link goes to the original either way.

SwiftembeddinggemmaSource GitHub
ARahim3/DigUp

1.9k stars138 forks

MITSwiftpushed today

Location, estimated · Not stated

What we saw

“Search your Mac by what's in your files: photos, PDFs, recordings, videos and code. On-device, built on Google DeepMind's Embed…”

Published 2026-10-08

Fetched 2026-10-12 23:01 UTC · 71b9b63aa131f592

What the account wrote — The account left it blank

Torn from — the README

JavaScriptSource GitHub
wheresryan22/anatomy

3.1k stars238 forks

MITJavaScriptpushed today

Location, estimated · Not stated

What we saw

“A Claude Code skill that turns a technical idea into a detailed, interactive isometric figure of a real machine. It makes SVG w…”

Published 2026-10-08

Fetched 2026-10-11 23:01 UTC · 30a09c855d09c4c2

What the account wrote — The account left it blank

Torn from — the README

GoSource GitHub
mhtsec/ARTEX

3.1k stars5.5k forks

AGPL-3.0Gopushed yesterday

Location, estimated · Not stated

What we saw

“AI 自主渗透测试系统 | 百度“agent+”攻防挑战赛冠军项目”

They claim — 点「更新」后:下载当前平台的发布包 → 比对 Release 的 SHA256SUMS → 用 -h 冒烟测试新二进制 → 暂存为 artex.

Published 2026-10-08

Fetched 2026-10-09 23:01 UTC · 0cc27db3d1a52caa

What the account wrote — The account left it blank

Torn from — the README

PythonaiSource GitHub
LosaLosSantos/aurelio-finance

2.2k stars286 forks

MITPythonpushed today

Location, estimated · Not stated

What we saw

“Open-source personal finance app with an AI financial advisor: track your net worth, investments, ETFs, cash and debts on your …”

Published 2026-10-06

Fetched 2026-10-09 23:01 UTC · 984d32f396b5debc

What the account wrote — The account left it blank

Torn from — the README

PythonSource GitHub
zhongerxin/iPhone-use

3.0k stars303 forks

MITPythonpushed today

Location, estimated · Not stated

What we saw

“让 Codex 通过 USB 操作真实 iPhone:引导安装、App 自动化、实时屏幕与截图回退。”

Published 2026-10-06

Fetched 2026-10-09 23:01 UTC · bdb641bc7307d99d

What the account wrote — The account left it blank

Torn from — the README

Goclaude-codeSource GitHub
yetone/magpie

1.3k stars75 forks

MITGopushed today

Location, estimated · China

What we saw

“Every agent's model. One place. Codex on DeepSeek, Claude Code on Kimi, from the menu bar.”

Published 2026-09-23

Fetched 2026-09-26 23:01 UTC · 7abcc96641ec3723

What the account wrote — China

Torn from — the README

JavaScriptSource GitHub
JohnHeibel/PDoomVideo

1.3k stars139 forks

no licenseJavaScriptpushed today

Location, estimated · United States

What we saw

“Source code for the Claude Opus 5.5 music video for I'm Upping My P(doom)”

Published 2026-09-22

Fetched 2026-09-25 23:01 UTC · 683af84eb2b0acdc

What the account wrote — Berkeley, California

Torn from — the README

MarkdownaiSource GitHub
pallavi-shekhar/ai-engineering-interview-questions-company-wise

1.5k stars144 forks

Apache-2.0Markdownpushed 6d ago

Location, estimated · India

What we saw

“Your Cheat Sheet For AI Engineering Interviews at Top AI Companies - Questions and Answers.”

They claim — How would you cut LLM serving cost by 10x?

Published 2026-09-19

Fetched 2026-09-25 23:01 UTC · 6d945172946e0f37

What the account wrote — Bangalore, India

Torn from — the README

C++apple-siliconSource GitHub
incoai/splash

915 stars92 forks

Apache-2.0C++pushed today

Location, estimated · Not stated

What we saw

“A local inference engine for Apple silicon, built around the model.”

They claim — 24 GB Macs can use smaller GGUF variants.

Published 2026-09-18

Fetched 2026-09-26 23:01 UTC · 92d2f0a31d978b91

What the account wrote — The account left it blank

Torn from — the README

PythonSource GitHub
kerpopule/hermes-jev-skills

875 stars83 forks

MITPythonpushed today

Location, estimated · Not stated

What we saw

“Jev-powered model routing, memory, compaction, skill selection, computer and browser use for Hermes agents (also Claude Code an…”

They claim — Mailbox sorting (jev mail): the subject and up to 2,500 characters of the body, redacted; the sender's domain (never the mailbox); a local class read off the address alone (automated for …

Published 2026-09-18

Fetched 2026-09-26 23:01 UTC · 1d5e91c79063d0be

What the account wrote — The account left it blank

Torn from — the README

PythonSource GitHub
ai-sucks-butt/ai-sucks-butt

1.7k stars0 forks

no licensePythonpushed today

Location, estimated · Not stated

What we saw

“If you think AI sucks, star the repo.”

Published 2026-09-14

Fetched 2026-09-16 23:01 UTC · 293a5d397115ca6c

What the account wrote — The account left it blank

Torn from — the README

HTMLSource GitHub
yifanzhang-pro/recurrent-looped-tranformer

858 stars84 forks

Apache-2.0HTMLpushed 2d ago

Location, estimated · Not stated

What we saw

“Official Project Page for Recurrent Looped Transformer (RLT)”

They claim — Training length is 32 operations; evaluation extends to 128 operations (4× the training length), with 2,048 test programs per task and length.

Published 2026-09-12

Fetched 2026-09-15 23:01 UTC · 675776a9a37fe26a

What the account wrote — The account left it blank

Torn from — the README

PythonaiSource GitHub
kruzovic7/ai-data-extractor

825 stars132 forks

MITPythonpushed 4d ago

Location, estimated · Not stated

What we saw

“Free open-source extractor for AI coding assistant chat histories. Supports Claude Code, Cursor, Windsurf, Aider, Cline/Roo Cod…”

Published 2026-09-11

Fetched 2026-09-15 23:01 UTC · 32cd43375b85b5e6

What the account wrote — The account left it blank

Torn from — the README

PythonaiSource GitHub
crwdla/tokentab

1.1k stars215 forks

MITPythonpushed 6d ago

Location, estimated · Unrecognised

What we saw

“A CLI that reads Claude Code, Codex, and Gemini CLI session logs and works out how much they cost, by model, project, and day.”

They claim — Cache hit under ~80%, consistently - your context probably isn't stable between calls, or caching isn't switched on.

Published 2026-09-07

Fetched 2026-09-13 23:03 UTC · 285e008f28d1dc05

What the account wrote — Newcastle

Torn from — the README

Pythonai-upscalingSource GitHub
perseval-BLR/NeuralScreen

823 stars45 forks

no licensePythonpushed today

Location, estimated · Not stated

What we saw

“DLSS 5 NR on your whole Windows desktop, in real time with Boost, FG, etc. RTX 30/40/50, 12 languages, user presets, recording …”

They claim — Windows 10 and multi-GPU systems are experimental — v1.12 fixes adapter/output selection from user logs, not local hardware.

Published 2026-09-06

Fetched 2026-09-15 23:01 UTC · b6fda159d2186867

What the account wrote — The account left it blank

Torn from — the README

PythonaiSource GitHub
jtydhr88/screenwriting-skills

1.2k stars134 forks

MITPythonpushed today

Location, estimated · Canada

What we saw

“Professional agent skills for screenwriting, television writing and dramaturgy”

Published 2026-09-06

Fetched 2026-09-15 23:01 UTC · 23ae32d1655cbeed

What the account wrote — Toronto, Canada

Torn from — the README

TypeScriptagentsSource GitHub
shadcn-ui/lint

1.7k stars25 forks

MITTypeScriptpushed yesterday

Location, estimated · Not stated

What we saw

“An agent-first linter for Tailwind design systems. Write design system rules that agents can verify.”

They claim — 10% to 48% less than with rules alone.

Published 2026-09-02

Fetched 2026-09-15 23:01 UTC · 2bae6979048769f9

What the account wrote — The account left it blank

Torn from — the README

RustSource GitHub
faisalkindi/DLSS5oneclick

769 stars55 forks

MITRustpushed today

Location, estimated · Not stated

What we saw

“One-click setup of the leaked DLSS 5 neural-rendering build for any DX11/DX12 game on RTX 20–50, with or without DLSS. ReShad…”

They claim — exe) - the game exe is detected automatically (the folder and two levels below it are searched, so bin\x64dx12\ and Unreal Binaries\Win64\ layouts work; Unity crash handlers, Unreal helpe…

Published 2026-08-31

Fetched 2026-09-13 23:03 UTC · 38bc58445c3c8fae

What the account wrote — The account left it blank

Torn from — the README

SwiftSource GitHub
vinzdg/codenotch

1.1k stars180 forks

MITSwiftpushed today

Location, estimated · Not stated

What we saw

“A macOS app that pins usage limits from Claude Code, Cursor, Codex, and Antigravity to a screen edge.”

Published 2026-09-05

Fetched 2026-09-06 23:01 UTC · 70a320489a90af6d

What the account wrote — The account left it blank

Torn from — the README

PythonaiSource GitHub
pierrenade/short-video-generator-AI

1.2k stars157 forks

MITPythonpushed yesterday

Location, estimated · Europe

What we saw

“Free open-source project designed for turning youtube-viedos into viral short videos. Highlight detection, subtitles, translati…”

Published 2026-09-05

Fetched 2026-09-06 23:01 UTC · 06354c4b68f4ed99

What the account wrote — Paris

Torn from — the README

PythonSource GitHub
Rion-Wu-tech/wechat-intelligence-hub

1.9k stars2.5k forks

AGPL-3.0Pythonpushed yesterday

Location, estimated · Elsewhere

What we saw

“Local-first WeChat intelligence system with a read-only CLI, Codex skills, searchable chat history, daily briefings, follow-ups…”

Published 2026-09-04

Fetched 2026-09-06 23:01 UTC · 50d742bbbcd0f43f

What the account wrote — Canberra,Australia

Torn from — the README

TypeScriptSource GitHub
Albert-Weasker/niubigeo

2.2k stars67 forks

Apache-2.0TypeScriptpushed today

Location, estimated · Not stated

What we saw

“Open-source AI brand visibility and competitor reports”

Published 2026-09-03

Fetched 2026-09-07 01:32 UTC · 138e30df9c592905

What the account wrote — The account left it blank

Torn from — the README

PythonSource GitHub
tigerless-labs/agent-memory

612 stars35 forks

no licensePythonpushed today

Location, estimated · United States

What we saw

This source gives a title and a link, nothing to quote.

They claim — Requires Python 3.12 or higher and uv.

Published 2026-09-01

Fetched 2026-09-08 16:21 UTC · 4719c1df68611e38

What the account wrote — United States of America

Torn from — the README

Pythonagent-infrastructureSource GitHub
Human-Agent-Society/reef

774 stars48 forks

Apache-2.0Pythonpushed today

Location, estimated · Not stated

What we saw

“Continual learning infra for self-improving agents”

Published 2026-08-31

Fetched 2026-09-07 01:32 UTC · e535803553f886d7

What the account wrote — The account left it blank

Torn from — the README

Pythonai-upscalingSource GitHub
Merserk/dlss5-visual-enhancer

655 stars54 forks

MITPythonpushed today

Location, estimated · Not stated

What we saw

“DLSS 5 Neural Video & Image Enhancer with Frame Interpolation”

They claim — Frame Interpolation: single-video and batch-video processing with NVIDIA DLSS Frame Generation, selectable output rates from 23.976 to 480 FPS, and a three-second preview for a single sel…

Published 2026-08-30

Fetched 2026-09-07 01:32 UTC · 58b954054e451478

What the account wrote — The account left it blank

Torn from — the README

Shellai-seoSource GitHub
Ryze-AI-Adgent/open-seo-mcp-skills

660 stars11 forks

MITShellpushed today

Location, estimated · Not stated

What we saw

“Open-source SEO + GEO skills for Claude — keyword research, rank tracking, site audits, backlinks, competitor gaps, AI visibi…”

They claim — Look closer and most are a UI over the DataForSEO API: you either bring your own key and pay per request, or pay a hosted subscription plus a ~28% markup on every data call.

Published 2026-08-29

Fetched 2026-09-08 16:21 UTC · c7ee6c65bdba0676

What the account wrote — The account left it blank

Torn from — the README

Zigai-agentsSource GitHub
duty1g/x64dbg-mcp-server

1.3k stars129 forks

MITZigpushed yesterday

Location, estimated · Not stated

What we saw

“x64dbg-MCP Server is a native MCP (Model Context Protocol) plugin for x64dbg that exposes the debugger's full functionality ove…”

They claim — MCP-powered agentic reverse engineering for x64dbg.

Published 2026-08-22

Fetched 2026-08-25 10:14 UTC · c1e90fdc5334c245

What the account wrote — The account left it blank

Torn from — the README

Pythonai-videoSource GitHub
cclank/lanshu-create-ai-presenter-video

883 stars153 forks

MITPythonpushed 4d ago

Location, estimated · Not stated

What we saw

“Provider-neutral Codex Skill for producing verified AI presenter videos from a script and an authorized presenter image.”

Published 2026-08-20

Fetched 2026-08-25 10:14 UTC · f2ced8b6406f4329

What the account wrote — The account left it blank

Torn from — the README

Pythona2aSource GitHub
wang2122/sprix-sage-router

2.0k stars24 forks

MITPythonpushed 4d ago

Location, estimated · Not stated

What we saw

“Sprix AI at 屿智同行 — state-aware SELF/COLLABORATE/HANDOFF routing for A2A agent networks.”

They claim — RouteLLM: Learning to Route LLMs with Preference Data, ICLR 2025 — preference-based cost-quality routing.

Published 2026-08-18

Fetched 2026-08-25 10:14 UTC · 1c90000bebc39c3d

What the account wrote — The account left it blank

Torn from — the README

JavaScriptagent-orchestrationSource GitHub
Spielewoy/autoprompt-skill

845 stars54 forks

MITJavaScriptpushed today

Location, estimated · Unrecognised

What we saw

“Autoprompt is a coding-agent skill that cuts failures by 45% on agentic coding tasks.”

They claim — Autoprompt is a coding-agent skill that cuts failures by 45% on agentic coding tasks.

Published 2026-08-17

Fetched 2026-08-25 10:14 UTC · 5d07e174c6aaeff9

What the account wrote — DE

Torn from — the README

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.