Douglas Galdino dougg0k

AI - Fine Tuning (Prompt Engineering)

I have been building a way to improve AI Chats and fix issues that I found while using it.

From what I understood, each AI chat (chatgpt, claude, gemini) may have their own issues, they may require different prompt engineering fixes. Given the effectiveness of each model, some phrases / constraints may not be needed in some chats.

Repeat the constraints in the conversation itself, not just in settings. This is important and should be the last one, since it can enforce the usage of all the constraints set, it can help prevent AI from ignoring them.

Order in defining the constraints, matter. Later constraints refine or specialize earlier ones. Lines below the other are preferred, rather than in a continued line. Small redudancy is not always bad, as it can reinforce the probability of use.

Obfuscate code helps.

JS Bundles
- https://github.com/vesselsoft/metro-minify-obfuscator
- https://github.com/jscrambler/jscrambler/tree/master/packages/jscrambler-metro-plugin
- https://github.com/MichaelXF/js-confuser - Adding this here as a note, because it could end up better than the javascript-obfuscator, but it might require a RN / Metro implementation with a working sourcemap 1 / 2 support.
- If using RePack, simply enable HermesBytecodePlugin.
Android
- Proguard / R8
iOS
Before deprecation, BITCODE was used. However there is a commercial one called iXGuard.

Edgerouter / EdgeOS / DoH / Security Settings

Notes

It seems that every firmware update, whatever was done to get nextdns into the router, will be gone, so you will have to redo the steps or at least some, if you confirm that it has happened.
Make sure to have NTP enabled and configured. To avoid bugs.
You can do through the UI CLI or using SSH, if you have enabled it and set a port.
You can delete any configuration done, by using delete, e.g., delete firewall all-ping.
These steps were confirmed to be working after a complete hardware reset.

In the screen with the Install button, do not press it.

Press Shift+F10 to open cmd
Type the following commands

REG ADD "HKLM\SYSTEM\Setup\LabConfig" /v BypassTPMCheck /t REG_DWORD /d 1 /f
REG ADD "HKLM\SYSTEM\Setup\LabConfig" /v BypassSecureBootCheck /t REG_DWORD /d 1 /f
REG ADD "HKLM\SYSTEM\Setup\LabConfig" /v BypassRAMCheck /t REG_DWORD /d 1 /f
REG ADD "HKLM\SYSTEM\Setup\LabConfig" /v BypassStorageCheck /t REG_DWORD /d 1 /f
REG ADD "HKLM\SYSTEM\Setup\LabConfig" /v BypassCPUCheck /t REG_DWORD /d 1 /f

	# Never use more than one --enable-features and --disable-features as it can overwrite the previous.
	# This was built using Arch Linux (Wayland) if some flags / features are not compatible with other OS, up to whoever uses, to remove them.

	# There are two threads that if ever get implemented into vscode, could bring improvements to the usage experience.
	# https://github.com/microsoft/vscode/issues/27378
	# https://github.com/microsoft/vscode/issues/221145

	# CPU
	--enable-features=RawDraw,HighEfficiencyModeAvailable,WebRTCPipeWireCapturer,FractionalScaleV1,WaylandPerSurfaceScale,NetworkServiceSandbox,FontAccess
	--disable-features=CalculateNativeWinOcclusion,BackForwardCache,NetworkServiceInProcess,VizDisplayCompositor,UseSkiaRenderer

	const puppeteer = require("puppeteer");
	const Xvfb = require("xvfb");

	async function getDataFromProtectedUrl({ url, operationName, query, variables }) {
	const xvfb = new Xvfb({
	silent: true,
	xvfb_args: ["-screen", "0", "1280x720x24", "-ac"],
	});
	xvfb.start((err) => {
	if (err) console.error(err);

	import got from "got";
	import tunnel from "tunnel";

	const PROXY_LIST = [
	{ ip: "", port: 80 },
	{ ip: "", port: 80 },
	{ ip: "", port: 80 },
	{ ip: "", port: 80 },
	];

	const getCountryFlagEmoji = countryCode => String.fromCodePoint(...[...countryCode.toUpperCase()].map(x => 0x1f1a5+x.charCodeAt(0)))

	getFlagEmoji('GB') // 🇬🇧
	getFlagEmoji('JP') // 🇯🇵
	getFlagEmoji('ZA') // 🇿🇦

	function App() {
	return (
	<RadioButton
	name="type"
	options={OPTIONS}
	defaultValue={"Initial Value"}
	isLoading={!!"Initial Value already loaded"}
	/>
	)
	}

	1- Create Github account.
	2- Create your private repo and push your files.
	3- Login into Netlify with Github and add your repo there.
	4- Given that you have your netlify.toml in your project already configured. Keep Post Process disabled.
	5- Add your domain to Netlify, but do not configure the DNS nor Alias.
	6- Create a Cloudflare account.
	7- Add your domain there and configure the DNS to theirs where you bought your domain.
	8- Configure the DNS adding CNAME @ with generated Netlify URL from your project in target and CNAME www with @ in target. Both with orange cloud activated (proxified).
	9- SSL > Full.
	10- Speed > Optimization, Activate Brotli and Rocket Loader and Minify all 3 options.