tairov

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.

tairov

cat <<EOF | kubectl create -f -
apiVersion: apps/v1
kind: Deployment
metadata:
  name: curl
spec:
  replicas: 1
  selector:
    matchLabels:
      app: curl

tairov

# Start pod named "sq-shell".

$ kubectl run sq-shell --image ghcr.io/neilotoole/sq

# Shell into the pod.

$ kubectl exec -it sq-shell -- zsh

tairov

# ffmpeg -i "$1" -vcodec h264 -threads 10 "$2"

process_mov_files() {
  {
    find . -name "*.mov" -print0 | while read -d $'\0' file
    do
      {
        NEW_NAME="$(echo "${file%.mov}" | sed 's/[ _]/-/g' | tr '[:upper:]' '[:lower:]')"
        ffmpeg -i "$file" -vcodec h264 -threads 10 -f mp4 "${NEW_NAME}.mp4"
      } < /dev/null

tairov

while true; do printf 'HTTP/1.1 200 OK\n\n%s' "pong $(date)" | nc -l 8888; done

tairov

 * To use Docker, the Docker daemon must be running as root. To automatically
 * start the Docker daemon at boot, add Docker to the default runlevel:
 *   rc-update add docker default
 * Similarly for systemd:
 *   systemctl enable docker.service
 *
 * To use Docker as a non-root user, add yourself to the 'docker' group:
 *   usermod -aG docker youruser

tairov

#!/usr/bin/env bash

# INPUT_SRC=input.mp4
INPUT_SRC=input.mp4
START_TIME=00:39:51
END_TIME=00:40:29
echo Start processing
ffmpeg -ss $START_TIME -i $INPUT_SRC -acodec copy -vcodec copy -c copy -to $END_TIME output2.mp4

tairov

INPUT_SRC=$(youtube-dl -f best -g 'https://www.youtube.com/watch?v=wnHW6o8WMas')
ffmpeg -ss 00:01:22 -i $INPUT_SRC -vframes 1 -q:v 2 output.jpg

tairov

# shared library

// mylib/vars/myGitlabStage.groovy
call (String _stage, Closure _closure, String _label = 'master') {
stage(_stage) {
  node (_label) {
    gitlabCommitStatus(name: _stage ) {
      _closure()
    }
  } // node

tairov

[main]     "stdout_lines": [
[main]         "JENKINS_HOME=/opt/jervis/../my_jenkins_home", 
[main]         "GIT_PREVIOUS_SUCCESSFUL_COMMIT=2287deab5c65f0597650ec7f83f779883ae4c5a0", 
[main]         "ENVIRONMENT_NAME=dev", 
[main]         "MAIL=/var/mail/jenkins", 
[main]         "USER=jenkins", 
[main]         "RUN_CHANGES_DISPLAY_URL=http://localhost:8080/page=changes", 
[main]         "HUDSON_URL=http://localhost:8080/", 
[main]         "NODE_LABELS=ansible docker docker-local-dace45f6a860 linux stable", 
[main]         "GIT_COMMIT=c690358ed6678d11fe96684690ab00c103c218c3",