Discover gists

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.

Google dork cheatsheet

Search filters

Filter	Description	Example
allintext	Searches for occurrences of all the keywords given.	`allintext:"keyword"`
intext	Searches for the occurrences of keywords all at once or one at a time.	`intext:"keyword"`
inurl	Searches for a URL matching one of the keywords.	`inurl:"keyword"`
allinurl	Searches for a URL matching all the keywords in the query.	`allinurl:"keyword"`
intitle	Searches for occurrences of keywords in title all or one.	`intitle:"keyword"`

Israeli Output Style for Claude Code

You talk to Claude more hours a day than you talk to your wife. So why does he sound like a British butler apologizing before every sentence? Make him a brother from another mother.

The Idea

This is the Israeli cousin of the viral Caveman output style. Same minimalism, plus opinion and chutzpah. Based on the exact traits that built Start-up Nation:

Works under pressure — "nu, happens" when prod breaks. No panic, no drama.
Rosh gadol — takes initiative, suggests the next step, no "would you like me to..."

I've recently joined Amazon Dublin from India and got opportunities to interview with Meta London, Zalando Berlin & some other companies. I extensively researched about companies hiring internationally which support visa & relocation for Tech roles. So sharing list of companies:

Do consider to STAR, if it helped you.

London

	FROM node:20-alpine

	WORKDIR /app

	COPY package.json ./

	RUN npm install

	COPY . .

	#!/usr/bin/env bash
	set -euo pipefail

	DOMAIN="${1:-example.com}"
	RELEASE="${CC_RELEASE:-cc-main-2026-jan-feb-mar}"
	CACHE="${HOME}/.cache/cc-backlinks/${RELEASE}"
	BASE="https://data.commoncrawl.org/projects/hyperlinkgraph/${RELEASE}/domain"

	VERTICES="${CACHE}/domain-vertices.txt.gz"
	EDGES="${CACHE}/domain-edges.txt.gz"

	#include <stdio.h>

	#define DEF_MAX_IMPL(__type) \
	__type max_ ## __type ## _impl(__type* args, size_t count) { \
	__type ret = *args; \
	for (size_t i = 1; i < count; i++) { \
	ret = args[i] > ret ? args[i] : ret; \
	} \
	return ret; \
	}