Sendhil Kumar R codingarchitect

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.

	#!/bin/bash
	#
	# Generates client and server certificates used to enable HTTPS
	# remote authentication to a Docker daemon.
	#
	# See http://docs.docker.com/articles/https/
	#
	# To start the Docker Daemon:
	#
	# sudo docker -d \

	If you'd like to use a .jar file in your project, but it's not available in any Maven repository,
	you can get around this by creating your own local repository. This is done as follows:

	1 - To configure the local repository, add the following section to your pom.xml (inside the <project> tag):

	<repositories>
	<repository>
	<id>in-project</id>
	<name>In Project Repo</name>
	<url>file://${project.basedir}/libs</url>

	#!/usr/bin/env bash
	# Convert your asciinema cast into a gif

	# https://github.com/asciinema/asciicast2gif#docker-image
	# https://brunswyck.blog/2017/05/17/record-your-bash-shell-and-upload-as-a-gif-file/

	# usage: in directory containing .cast file and this script: ./asciigifs <file name without .cast extension>

	command -v docker >/dev/null 2>&1 \|\| (echo "docker must be installed" && exit 1)