Pavel paveljurca

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.

Manage Windows Servers using Ansible

This is a real-time demo of how to set up your windows servers so they can be managed by ansible. I've created a YouTube companion video demo here: https://youtu.be/aPN18jLRkJI

Subscribe To Me On YouTube: https://bit.ly/lon_sub

The following is covered by this demo:

Setup a windows 2016 (or 2019/2021) server so ansible can manage it (over HTTPS)
Setup my Centos 8 server so it can manage windows servers using winrm

As a flat file CMS, Kirby stores all content in folders and files. This is not only true for pages, but also for files and user accounts, which are by default stored in the /site/accounts folder.

But in the same way as you can incorporate content from other sources like spreadsheets, feeds or databases to create (link: text: virtual pages in Kirby), you can also replace users stored in the file system with users stored in a database (or some other source).

In this recipe, we will see how we can achieve this. For this purpose, we will start with a simple read-only solution and then extend this basic setup step-by-step into a solution that lets us create and edit database users from the Panel, including multi-language user content.

Note that using users from a database as described in this recipe will **not** solve potential performance issues with thousands of users.

Alpine.js fetch data on `x-init`

A Pen by Marcus Obst on CodePen.

License.

Conventional Commit Messages

See how a minor change to your commit message style can make a difference.

git commit -m"<type>(<optional scope>): <description>" \
  -m"<optional body>" \
  -m"<optional footer>"

	# Copyright 2021 Google LLC

	# Licensed under the Apache License, Version 2.0 (the "License");
	# you may not use this file except in compliance with the License.
	# You may obtain a copy of the License at

	# https://www.apache.org/licenses/LICENSE-2.0

	# Unless required by applicable law or agreed to in writing, software
	# distributed under the License is distributed on an "AS IS" BASIS,

	#!/usr/bin/env python3
	#
	# Convert traceroute command output to a graphviz DOT graph
	# Use `jc` to convert `traceroute` output to JSON before piping to tr2dot.py
	#
	# Version 1.0
	# Kelly Brazil (kellyjonbrazil@gmail.com)
	#
	# Requires graphviz (pip3 install graphviz)
	#

	const echoPostRequest = {
	url: 'https://<my url>.auth0.com/oauth/token',
	method: 'POST',
	header: 'Content-Type:application/json',
	body: {
	mode: 'application/json',
	raw: JSON.stringify(
	{
	client_id:'<your client ID>',
	client_secret:'<your client secret>',

	/* CSS Custom Properties */

	:root {
	--font-family-default: 'Merriweather', serif;
	--font-family-alt: 'Merriweather Sans', sans-serif;

	--font-color-default: #b1a18;
	--font-color-alt: #f5f5f3;

	--font-link-default: #a62339;

	/* CSS Custom Properties */

	:root {
	--font-family: 'Georgia', serif;
	--font-family-alt: 'Helvetica', Arial, sans-serif;
	--font-weight: 400;
	--font-weight-bold: 700;
	--font-weight-black: 900;

	/* 3:4 perfect fourth scale */