Charles Wei JavaCS3

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.

	<!DOCTYPE html>
	<html>
	<head>
	<meta charset="utf-8">
	<title>JS Bin</title>
	<script src="http://ajaxorg.github.io/ace-builds/src/ace.js"></script>
	<style>
	#editor { position: absolute; top: 0; left: 0; right: 0; bottom: 0;}
	.ace_link_marker {
	position: absolute;

	///
	/// Simple pooling for Unity.
	/// Author: Martin "quill18" Glaude (quill18@quill18.com)
	/// Latest Version: https://gist.github.com/quill18/5a7cfffae68892621267
	/// License: CC0 (http://creativecommons.org/publicdomain/zero/1.0/)
	/// UPDATES:
	/// 2015-04-16: Changed Pool to use a Stack generic.
	///
	/// Usage:
	///

	黄浦区
	街道南京东路街道外滩街道半淞园路街道小东门街道豫园街道老西门街道

	卢湾区
	街道五里桥街道打浦桥街道淮海中路街道瑞金二路街道

	徐汇区
	街道天平路街道湖南路街道斜土路街道枫林路街道长桥街道田林街道虹梅路街道康健新村街道徐家汇街道凌云路街道龙华街道漕河泾街道
	镇华泾镇
	类似乡级单位漕河泾新兴技术开发区

	import subprocess
	import select
	from logging import DEBUG, ERROR


	def call(popenargs, logger, stdout_log_level=DEBUG, stderr_log_level=ERROR, **kwargs):
	"""
	Variant of subprocess.call that accepts a logger instead of stdout/stderr,
	and logs stdout messages via logger.debug and stderr messages via
	logger.error.