Junliang Hu vtta

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.

Comparison of ZSH frameworks and plugin managers

Changelog

update 1: add a FAQ section
update 2: benchmark chart and feature comparison table
update 3:
- improve the table with missing features for antigen
- new zplg times result

TLDR

Generic makefile example for a C++ project

If you need a small makefile introduction/reference you can take a look at my notes https://www.mauriciopoppe.com/notes/os/bin/make/

Project structure

. project
├── Makefile
├── build

注意：本文内容适用于 Tmux 2.3 及以上的版本，但是绝大部分的特性低版本也都适用，鼠标支持、VI 模式、插件管理在低版本可能会与本文不兼容。

Tmux 快捷键 & 速查表 & 简明教程

启动新会话：

tmux [new -s 会话名 -n 窗口名]

恢复会话：

	import hashlib
	import struct
	import sqlite3

	def md5hash(buf):
	return hashlib.md5(buf.encode("utf-16")[2:]).digest()

	def md5cmp(buf, postfix, a1, a2, a3, a4):
	if len(buf) < postfix:
	return False

	/*
	* Linux x86(-64) - execve("/bin/sh", ["/bin/sh", 0], 0) shellcode (38 bytes)
	* 31c050488b1424eb105478065e5fb03b0f05595b40b00bcd80e8ebffffff2f62696e2f736800
	*
	* - offset - bytes 32-bit code 64-bit code
	* 0x00000000 31c0 xor eax, eax xor eax, eax
	* 0x00000002 50 push eax push rax
	* 0x00000003 48 dec eax rex.w
	* 0x00000004 8b1424 mov edx, dword [esp] mov rdx, qword [rsp]
	* 0x00000007 eb10 jmp 0x19 jmp 0x19

	#!/bin/bash

	# https://tug.org/mactex/uninstalling.html

	echo "# Uninstall TeX"
	echo "rm -rf /usr/local/texlive"

	echo "# Uninstall GUI Applications"
	echo "rm -rf /Applications/TeX"

	<llppconfig>
	<ui-font size='18'>
	<![CDATA[/home/earnest/.local/share/fonts/Fontin-Regular.otf]]>
	</ui-font>

	<defaults zoom='100'
	auto-scroll-step='12'
	horizontal-scroll-step='24'
	case-insensitive-search='true'
	uri-launcher='$BROWSER "%s"'

	#### Example SSH config file

	# Host = a list of domains, IPs and personalized aliases that use this config
	# entry. These alias names follow "Host" and are space delimited.
	# e.g. Host 192.168.1.1 example.com example1 myalias1
	# Hostname = the IP or real hostname
	# e.g. Hostname 192.168.1.1
	# User = the username
	# e.g. User jdoe
	# Port = if this is a non standard port, i.e. not 22

	#
	# STL GDB evaluators/views/utilities - 1.03
	#
	# The new GDB commands:
	# are entirely non instrumental
	# do not depend on any "inline"(s) - e.g. size(), [], etc
	# are extremely tolerant to debugger settings
	#
	# This file should be "included" in .gdbinit as following:
	# source stl-views.gdb or just paste it into your .gdbinit file