roktas

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.

roktas

" ----------------------------------------------------------------------------------------------------------------------
" Settings
" ----------------------------------------------------------------------------------------------------------------------

set confirm=permdelete
set history=500
set hlsearch
set ignorecase
set incsearch
set nofollowlinks

roktas

local wezterm = require("wezterm")

------------------------------------------------------------------------------------------------------------------------
-- Constants and helpers
------------------------------------------------------------------------------------------------------------------------

local FONT_FAMILY = wezterm.font("Spleen 32x64")
local FONT_SIZE = 16
local CHAR_SOLID_LEFT = wezterm.nerdfonts.pl_right_hard_divider
local COLOR_BACKGROUND = "#2e3440"

roktas

Ekran görüntüleri

• VSCode CodeSnap Extension

• "Operator Mono Book", 18pt https://github.com/keyding/Operator-Mono

• Light Modern ve Github Light

• background: #fff

roktas

# frozen_string_literal: true

require "sevgi/sundries/printer"

module Sevgi
  module Graphics
    module Mixtures
      module Print
        EXT = ".svg"

roktas

#!/usr/bin/env bash

declare -gr ROOT=~/Dropbox/src/strokes

alef() {
	local P

	echo >&2 "alef"

	for P in g i; do

roktas

# frozen_string_literal: true

module Base
  module Header
    def call(page)
      n     = 3
      width = page.grid.width / n.to_f

      rect id: "Header-Cell-Left",   x: 0,         y: 0, width:, height: 9, class: %w[ header ]
      rect id: "Header-Cell-Center", x: width,     y: 0, width:, height: 9, class: %w[ header ]

roktas

Sözdizimi

• Sorular bir soru bir dosya olabileceği gibi, bir dosyada birden fazla soru olabilir

• Dosyada sorular === ile ayrılır

• Grup sorularında alt sorular --- ile ayrılır

• Şıklar A) şeklinde

roktas

roktas

#!/usr/bin/env -S ruby -S sevgi

SVG :inkscape, Sevgi::Graphics::Canvas.new(width: 150, height: 30) do
  rect x: 0, y: 0, width: 150, height: 30, stroke: "none", fill: "maroon"
  text "TEXT", x: "50%", y: "50%", "dominant-baseline": "middle", "text-anchor": "middle", fill: "white"
end.Out