oldcai

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.

oldcai

/* Useful celery config.

app = Celery('tasks',
             broker='redis://localhost:6379',
             backend='redis://localhost:6379')

app.conf.update(
    CELERY_TASK_RESULT_EXPIRES=3600,
    CELERY_QUEUES=(
        Queue('default', routing_key='tasks.#'),

oldcai

Lecture 1: Introduction to Research — [📝Lecture Notebooks] [▶️Video]
Lecture 2: Introduction to Python — [📝Lecture Notebooks] [▶️Video]
Lecture 3: Introduction to NumPy — [📝Lecture Notebooks] [▶️Video]
Lecture 4: Introduction to pandas — [📝Lecture Notebooks] [▶️Video]
Lecture 5: Plotting Data — [📝Lecture Notebooks] [[▶️Vide

oldcai

#!/bin/bash
version='stable'
#version='latest'
mkdir -p /var/backups/rancher
cd /var/backups/rancher
container_id=$(docker ps|grep rancher:${version}|awk '{ print $1 }')
echo $container_id
docker stop $container_id
today=$(date +%Y%m%d)
docker create --volumes-from $container_id --name rancher-data-$today rancher/rancher:${version}

oldcai

// ==UserScript==
// @name         copy 500px images
// @namespace    http://www.oldcai.com/
// @version      0.1
// @description  copy 500px images
// @author       You
// @match        https://500px.com/photo/*
// @grant        none
// ==/UserScript==

oldcai

#!/usr/bin/env python
import os
import sys

from gevent import monkey
from psycogreen.gevent import patch_psycopg

# The threading module raises a KeyError if already loaded
# when monkey patching it.
if "threading" in sys.modules:

oldcai

#!/bin/sh
/jffs/scripts/xunlei stop
/jffs/scripts/swap stop
/jffs/scripts/remount start
/jffs/scripts/swap start
/jffs/scripts/xunlei start

oldcai

Debian

cd /tmp
# 下载源码
git clone https://github.com/madeye/shadowsocks-libev.git
# 开始编译
cd shadowsocks-libev
./autogen.sh
./configure --prefix=/usr && make
make install

oldcai

[Unit]
Description=Great Shadowsocks Server
Wants=network-online.target
After=network.target

[Service]
Type=simple
PIDFile=/run/shadowsocks.pid
ExecStart=/usr/local/bin/ss-server -c /etc/sysconfig/shadowsocks

oldcai

<?xml version="1.0"?>
<root>
	<item>
		<name>Change KeyPad 0 to Num 0</name>
		<identifier>private.change_keypad0_to_num0</identifier>
		<autogen>__KeyToKey__ KeyCode::KEYPAD_0, KeyCode::KEY_0</autogen>
	</item>
	<item>
		<name>Change KeyPad 1 to Num 1</name>
		<identifier>private.change_keypad1_to_num1</identifier>