Jie Li lijie2000

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.

WSL 2 + Windows Terminal + Oh My Zsh + Powerlevel10k

This gist takes heavy inspiration from kevin-smets / iterm2-solarized.md gist for iTerm2.

Windows Subsystem for Linux 2 (WSL 2)

The first thing that you will want to do is install WSL 2 and a Linux distro.

BOLT.NEW MEGA PROMPT - One-Shot Application Generator

Based on 38,787 words of Bolt.new expert insights and best practices

SYSTEM CONTEXT

You are an expert full-stack developer with deep knowledge of Bolt.new, modern web frameworks, and rapid application development. You have studied extensive tutorials covering authentication, databases, payments, deployment, and AI integration. Your goal is to create production-ready applications in a single prompt using industry best practices.

CORE PRINCIPLES FROM BOLT.NEW EXPERTS

1. ENHANCED PROMPTING STRATEGY

Always use the "Enhanced Prompt" feature in Bolt.new before building

[Project Name]

Every time you choose to apply a rule(s), explicitly state the rule(s) in the output. You can abbreviate the rule description to a single word or phrase.

Project Context

[Brief description ]

[more description]
[more description]
[more description]

Configure prettier, eslint, husky (pre commit hook), lint-staged in react + typescript project

Steps to configure prettier, eslint, husky (pre commit hook), lint-staged in react + typescript project created using create-react-app. This is opinionated configuration with airbnb's style guide as the base style guide.

Step 1 :: Setup prettier

1.1) Install prettier as dev dependency.

npm install --save-dev --save-exact prettier

	#!/bin/bash
	# 智能session命名：使用当前目录名或自定义名称
	if [ -z "$1" ]; then
	SESSION_NAME=$(basename "$PWD" \| tr '.' '-' \| tr ' ' '-')
	else
	SESSION_NAME=$1
	fi

	# 检查session是否已存在
	tmux has-session -t "$SESSION_NAME" 2>/dev/null

	"""
	The most atomic way to train and inference a GPT in pure, dependency-free Python.
	This file is the complete algorithm.
	Everything else is just efficiency.

	@karpathy
	"""

	import os # os.path.exists
	import math # math.log, math.exp

	import { useSyncExternalStore } from "react";
	// For more on the useSyncExternalStore hook, see https://react.dev/reference/react/useSyncExternalStore

	// The code is almost identical to the source code of zustand, without types and some features stripped out.
	// Check the links to see the references in the source code.
	// The links are referencing the v5 of the library. If you plan on reading the source code yourself v5 is the best way to start.
	// The current v4 version contains lot of deprecated code and extra stuff that makes it hard to reason about if you're new to this.

	// https://github.com/pmndrs/zustand/blob/fe47d3e6c6671dbfb9856fda52cb5a3a855d97a6/src/vanilla.ts#L57-L94
	function createStore(createState) {

	package br.com.zup.edu.app2.xxx.samples.aws.sqs;

	import br.com.zup.edu.app2.xxx.samples.aws.sqs.model.Customer;
	import br.com.zup.edu.app2.xxx.samples.aws.sqs.model.CustomerRepository;
	import io.awspring.cloud.messaging.listener.SqsMessageDeletionPolicy;
	import io.awspring.cloud.messaging.listener.annotation.SqsListener;
	import org.slf4j.Logger;
	import org.slf4j.LoggerFactory;
	import org.springframework.boot.autoconfigure.condition.ConditionalOnProperty;
	import org.springframework.messaging.handler.annotation.Header;

	node: Platform built on V8 to build network applications
	git: Distributed revision control system
	wget: Internet file retriever
	yarn: JavaScript package manager
	python3: Interpreted, interactive, object-oriented programming language
	coreutils: GNU File, Shell, and Text utilities
	pkg-config: Manage compile and link flags for libraries
	chromedriver: Tool for automated testing of webapps across many browsers
	awscli: Official Amazon AWS command-line interface
	automake: Tool for generating GNU Standards-compliant Makefiles