Manassarn "Noom" Manoonchai narze

name	feel-lab
description	A catalogue of 41 perceived-performance and interface-craft techniques that make an app feel faster without changing how fast it actually is — skeletons, optimistic UI, spring physics, the Doherty threshold, and more. Invoke with /feel-lab.
disable-model-invocation	true

Feel Lab

A reference catalogue of 41 techniques for perceived performance. The thesis: nothing here makes an app faster; everything here makes it feel faster. Each technique has a real off-state — the point is always the contrast between "trick on" and the bare interaction underneath.

Bad Apple!! in a tldraw offline canvas

Plays the Bad Apple!! silhouette animation inside the tldraw offline desktop app as a durable document script. A custom baplayer shape renders each frame as inline SVG (decode-free), driven at 30 fps, with Play / Pause, Reset, and a click-to-seek progress bar. An optional local audio file is used as the master clock so the animation stays in sync with the music.

Files

llm-wiki-extended Architecture

This vault follows the llm-wiki-extended pattern — an extension of Karpathy's llm-wiki with PARA organization, idea inbox, action orientation, and bidirectional cross-project knowledge flow.

The Original Insight (Karpathy's llm-wiki)

The core idea: Instead of using RAG (retrieve-at-query-time), maintain a persistent, LLM-curated wiki that sits between raw sources and the user. When you add a source, the LLM doesn't just index it — it integrates it:

Reads the source completely

	# Usage: Start docker compose, then point Elastic APM to http://localhost:8200
	# Access Kibana at http://localhost:5601

	services:
	elasticsearch:
	image: docker.elastic.co/elasticsearch/elasticsearch:${ES_VERSION:-9.0.2}
	ports:
	- 9200:9200
	- 9300:9300
	environment:

	(function () {
	'use strict';
	const enabled = true;

	function log() {
	// Uncomment to debug
	// console.log(arguments)
	}

	function findNotificationBubble() {

	require "digest"

	# Cuid2 in Ruby, converted from cuid2.js via ChatGPT 4
	# https://github.com/paralleldrive/cuid2/blob/main/src/index.js
	# Example usage
	# cuid_instance = Cuid2.new
	# puts cuid_instance.generate
	# cuid_singleton = Cuid2.generate
	# puts cuid_singleton

	(async () => {
	async function getUsersInGeneralChannel(marker = null) {
	const response = await fetch(
	"https://edgeapi.slack.com/cache/TSK2FNWP3/users/list?fp=ed",
	{
	headers: {
	accept: "/",
	"accept-language": "en-US,en;q=0.9",
	"content-type": "text/plain;charset=UTF-8",
	"sec-ch-ua": '"Not)A;Brand";v="24", "Chromium";v="116"',

	function splitthai(str) {
	// https://github.com/nota/split-graphemes/blob/master/src/thai.js
	const letter = '[\\u0E00-\\u0E7F]';
	const trailingLetter = '[\\u0E31\\u0E33-\\u0E3A\\u0E47-\\u0E4E]';
	const thai = `${letter}${trailingLetter}*`;
	const splitter = new RegExp(`(${thai})`, 'gu');

	return str.replace(/ำ/g, 'ํา').replace(/แ/g, 'เเ').match(splitter) \|\| [];
	}