behumble

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.

behumble

public class ForLoop {
	public static void main(String[] args) {
		for(char c : "Hello") {
			System.out.println(c);
		}
	}
}

behumble

package com.itextpdf.sample;

import com.itextpdf.text.Document;
import com.itextpdf.text.Paragraph;
import com.itextpdf.text.pdf.PdfDestination;
import com.itextpdf.text.pdf.PdfOutline;
import com.itextpdf.text.pdf.PdfPTable;
import com.itextpdf.text.pdf.PdfWriter;

import java.io.FileOutputStream;

behumble

package com.itextpdf;

import com.itextpdf.text.Document;
import com.itextpdf.text.Paragraph;
import com.itextpdf.text.pdf.PdfDestination;
import com.itextpdf.text.pdf.PdfOutline;
import com.itextpdf.text.pdf.PdfPTable;
import com.itextpdf.text.pdf.PdfWriter;

import java.io.FileOutputStream;

behumble

Download Udemy course videos using youtube-dl

$ youtube-dl --list-extractors | grep udemy

Steps

1. Get link to the course to download. e.g. https://www.udemy.com/course-name/
2. Login into udemy website, save the cookie from chrome using Chrome (Cookie.txt)[1] export extension. Save it to file udemy-cookies.txt
3. Get the link of the video that you want to download. usually in format. Use the command provided below where you have to replace the {course_link} and {path_to_cookies_file} with respective paths.

# $ youtube-dl {course_link} --cookies {path_to_cookies_file}

behumble

Keybase proof

I hereby claim:

• I am behumble on github.
• I am behumble (https://keybase.io/behumble) on keybase.
• I have a public key whose fingerprint is 382D 1D88 8BB7 458B 0BCC CDDF 0775 028B CE4B EAD5

To claim this, I am signing this object:

behumble

#!/usr/bin/env python

import sys
import json
import csv
reload(sys)
sys.setdefaultencoding('utf-8')

if len(sys.argv)<2:
    print "json file name please"

behumble

#!/usr/bin/env python

import os
from datetime import date
from datetime import timedelta
from subprocess import call

a_week = timedelta(days=7)
a_week_ago = date.today() - a_week

behumble

#!/usr/bin/env groovy

String extractPackageName(File manifestFile) {
	return new XmlSlurper().parse(manifestFile).@package
}

void listDuplicates() {
	for( e in packageToFiles ) {
		files = e.value
		if(files.size>1) {

behumble

package com.thinkfree.touchspan;

import android.graphics.Canvas;
import android.graphics.Color;
import android.graphics.ColorFilter;
import android.graphics.Paint;
import android.graphics.PixelFormat;
import android.graphics.Rect;
import android.graphics.drawable.Drawable;