Skip to content

Instantly share code, notes, and snippets.

@scruel
scruel / fix-ms-input-pinyin-phrase.py
Last active October 8, 2026 09:14
修正微软拼音输入法无法添加多个格式化自定义短语的问题,默认添加 sj 和 rq 两个自定义短语
"""
修正微软拼音输入法无法添加多个格式化自定义短语的问题
Author: Scruel Tao
"""
import os
import re
import pathlib
import traceback
from pathlib import Path
@qwertychouskie
qwertychouskie / README.md
Last active October 8, 2026 09:13
"I Can't Believe It's Not GitHub" CSS theme for Forgejo (Codeberg, etc)

A CSS styleshoot for Forgejo that makes it look and feel a bit more like GitHub.

This started off with me just changing colors, but quickly grew to also adjusting various parts of the UI to be more GitHub-like as well. The most obvious one is probably the top bar contents being inlined (except for the home page), which saves a decent amount of screen real estate, but there's a ton of tweaks all across the UI to make it more in line with GitHub.

It definitely doesn't cover everything, nor is it perfect, but it's close enough for my use, and probably most people, though further suggestions/tweaks to the CSS are welcome.

I recommend setting Settings -> Appearance -> Default theme to gitea-auto, as that's what this was developed against. Most focus went into the dark theme, but it was sanity-checked in light mode as well.

I'm personally using the FireMonkey extension for Firefox to set the theme, but should work with any CSS changing addon in your browser of choice.

@qoomon
qoomon / conventional-commits-cheatsheet.md
Last active October 8, 2026 09:09
Conventional Commits Cheatsheet
@NARKOZ
NARKOZ / README.MOTHERFUCKING.md
Last active October 8, 2026 09:09
This gist lists Motherfucking websites. Plain HTML. No frameworks. Read them, learn, stop overengineering.

Motherfucking Websites

A concise list of the Motherfucking-style sites — because sometimes the web needs blunt, unapologetic clarity.

@joonan30
joonan30 / AGENTS.md.template
Last active October 8, 2026 09:06
LLM Wiki: building a compounding knowledge base of academic papers with AI agents (revised 2026-08)
# LLM Wiki — [YOUR FIELD]
A personal knowledge base of [YOUR FIELD] papers, following [Karpathy's LLM Wiki pattern](https://gist.github.com/karpathy/1dd0294ef9567971c1e4348a90d69285):
```
Original PDF → sources/*.md → wiki/{category}/*.md → wiki/overviews/ + wiki/concepts/
```
Language policy: all wiki content is in English. Conversation can be in any language.
<!DOCTYPE html>
<html lang="en">
<head>
<link rel="icon" type="image/x-icon" href="/favicon.ico">
<title>Rebane's Discord Colored Text Generator</title>
<meta charset="UTF-8">
<meta name="description" content="Rebane's Discord Colored Text Generator">
<meta name="author" content="rebane2001">
<meta content="#5865F2" name="theme-color">
<script id="discord:component-embed" type="application/json">
@as3eem
as3eem / HOPE Dataset
Created February 6, 2021 17:27
This repository contains youtube urls of therapy sessions conatinting conversations between therapist and patient over youtube. The corpus is named HOPE and is a part of research at IIIT Delhi, IND
https://youtu.be/ihDezGfSvQs
https://youtu.be/jKIFMKsPMas
https://youtu.be/9om2tedf9oo
https://youtu.be/FQOrzyBuAi8
https://youtu.be/88qr91P1lsA
https://youtu.be/Na11PJGFokQ
https://youtu.be/s6UW9b6zncc
https://youtu.be/QpIV2Zqf8F0
https://youtu.be/rmjBUBlFfU8
https://youtu.be/rGyA5IgmzuM
@myheart111
myheart111 / How to download from pan.baidu.com, pan.quark.cn and 123pan without an account (2026).md
Last active October 8, 2026 08:59
How to download from pan.baidu.com, pan.quark.cn and 123pan without an account (2026)

How to download from pan.baidu.com, pan.quark.cn and 123pan without an account (2026)

Looking for a way to download a Baidu Netdisk (百度网盘), Quark Pan (夸克网盘) or 123pan (123云盘) share link without a Chinese phone number, without the app and without the 50 KB/s free speed? This is the step-by-step I use in 2026. One website, three clouds, no account — you paste the link, you see the real size and the price, then you get a normal download link.

🔄 Update (25 September 2026) — 123pan added The tricks that used to work — PanDownload, baidu.kinh.cc, browser userscripts, "cracked" Netdisk clients, Telegram "parser" bots — are dead, banned within days, or malware. The same wall now exists on Quark Pan, which replaced Baidu Pan on most Chinese forums, and since 26 November 2025 on 123pan too: the "guest download under 100 MB" is gone, every download needs a logged-in account with a +86 number and a 10 GB monthly quota. The method below works for **pan.baidu.com/s/…, `pan.quark.cn/s/…

@tdcosta100
tdcosta100 / WSL2GUIWSLg-Wayland-en.md
Last active October 8, 2026 08:58
Using full desktop shell in WSL2 using WSLg (Wayland)

Full desktop shell in WSL2 using WSLg (Wayland)

Note

If you want a more complete experience and already tried the Xvnc tutorial, maybe you should try the WSLg (Xwayland) instead.

Warning

This is a work in progress tutorial. Things done here may break existing functionality, so be careful!

In this tutorial, we will install and use a full GNOME Desktop environment in WSL2, without any external software. The only requirement is a working WSLg installation. At the moment, the instructions are only for Ubuntu (20.04, 22.04 and 24.04) distros and GNOME, but you can request me to test other distros and desktop environments.

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.