Skip to content

Instantly share code, notes, and snippets.

View 0xdevalias's full-sized avatar
👀
Open to opportunities

Glenn 'devalias' Grant 0xdevalias

👀
Open to opportunities
View GitHub Profile

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.

@CappyT
CappyT / README.en.md
Created January 26, 2026 15:55
Help Keep Iran Connected: Conduit Proxy Setup

🗽 Conduit Proxy: Help Keep Iran Connected

TL;DR: Help the Iranian population bypass censorship by running a lightweight, low-maintenance proxy. It requires no public IP, acts only as a bridge (not an exit node), and can be deployed via Kubernetes or Docker in seconds.

ℹ️ Context

Access to information is a human right. Due to severe internet censorship in Iran, many citizens are cut off from the rest of the world.

import { detectReactBuildType, getRDTHook, type ReactRenderer } from "bippy";
const devRenderers: ReactRenderer[] = [];
const originalScheduleRefreshFns: ReactRenderer["scheduleRefresh"][] = [];
const originalScheduleUpdateFns: ReactRenderer["scheduleUpdate"][] = [];
const originalScheduleRootFns: ReactRenderer["scheduleRoot"][] = [];
let isHMRDisabled = false;
interface QueuedUpdate {
args: any[];
@programming086
programming086 / Links.md
Last active June 4, 2026 20:48
Reverse engineering macOS/iOS/iPadOS
@Cortexelus
Cortexelus / udio.py
Last active November 20, 2024 02:03
Automate Udio (2024-04-13)
# AUTOMATE UDIO
# by DADABOTS dadabots.com
# was working on 2024-04-13
# tldr; get your login cookie, use it to automate udio from python
import requests
import json
from time import sleep
import re
@m4ll0k
m4ll0k / _deobfuscating-unminifying-obfuscated-web-app-code.md
Created March 26, 2024 20:04 — forked from 0xdevalias/_deobfuscating-unminifying-obfuscated-web-app-code.md
Some notes and tools for reverse engineering / deobfuscating / unminifying obfuscated web app code
@ThevenRexOff
ThevenRexOff / _deobfuscating-unminifying-obfuscated-web-app-code.md
Created March 12, 2024 18:23 — forked from 0xdevalias/_deobfuscating-unminifying-obfuscated-web-app-code.md
Some notes and tools for reverse engineering / deobfuscating / unminifying obfuscated web app code
@d00m4ace
d00m4ace / ai-voice-cloning.md
Last active July 4, 2025 21:48
AI Voice Cloning
const fs = require("fs");
const { SourceMapConsumer, SourceMapGenerator } = require("source-map");
function reverseSourceMap(sourceMap) {
return SourceMapConsumer.with(sourceMap, null, (consumer) => {
const reversedMap = new SourceMapGenerator();
consumer.eachMapping((mapping) => {
reversedMap.addMapping({
generated: {
line: mapping.originalLine,
@ivanfioravanti
ivanfioravanti / sd_webp_workflow.json
Created November 27, 2023 06:56
(basic) ComfyUI Stable Video Diffusion workflow webp
{
"last_node_id": 25,
"last_link_id": 43,
"nodes": [
{
"id": 14,
"type": "VideoLinearCFGGuidance",
"pos": [
487.799932861328,
265.6999450683599