I built a personal closed source forward proxy server. A forward proxy makes connections to other systems on your behalf. It's as though the connection originated from the proxy server itself. This is incredibly useful for privacy, and to bypass some IP-level detection. I currently use this on peekabooo.

It supports HTTP, WS, SOCKS5 + UDP, HTTP-CONNECT, and If you ever need to use my forward proxy, send me a message.

You should expect a 407 or connection error on the proxy URL. Access is restricted to authorized users. This is intentional.

Vanity is a fast custom wallet generator that lets you get addresses with custom text at the beginning or end. It takes your set prefix or (and) suffix through stdin using readline, then uses all available CPU threads to grind keys at maximum speed until a match is found.

It currently supports the Solana, and EVM-chains. I'll add more chains as I get to understand them.

Peekabooo will get you screenshots of a url in an image format. For accessibility, I made it an open API, and nope...no need to install anything. You, me, anyone can access it. Just don't abuse it, for you will be rate-limited.

This API makes use of puppeteer to open a browser, load the site you set for screenshot, screenshot it, and streams it back to you.

Useful when you make previews of websites, without need to update previews when a new design or update is released. It is in use on this site.

In 2022, when ChatGPT was launched, I set out to build a chatbot, aiming for something spectacular, despite the limited info at the time. Syntaxx was designed to be an offline web chatbot for chatting directly in the browser, using indexedDB to locally store chat, allowing creation, deletion, and getting right back, without a single server involvement.

My approach was on flexible-static rules for replies, without context from previous messages, so it was nowhere near GPT-3's capabilities.

I paused to learn more, but realized soon, it was a bigger game, so I let it go. I could revamp it now for today's use, but I'm archiving it as one of my early attempt.

A react website, supposedly to be filled with quality elements, designed by open source contributions. I thought ahead with all the competitions out there, no way this was going to be a success. Oh well, I built what it would be like, with some actual functioning parts.

The goal was for devs from all around including me to contribute to the project, showcasing their abilities, taking the opportunity to get sponsored on their “Buy me a coffee” links, or on socials.

I also needed to do some API for the application, that devs could use to pull their designs to whatever site. But limited to frontend at the time, and no support to pay for server operations, I.....achieved it on FRONTEND.

Who does that? Me! Well, with the help of Vercel's free hosting and domain, and config, so anyone could iframe their design to their site.

A web-based 2D game built with Canvas. The idea came from a game whose source code a friend shared with me, challenging I recreate it from scratch. I did that, but this was no copy and paste. I only needed to understand the game’s rules, then wrote my own version for Zrang.

The game took me about 4 days to build despite not knowing anything about canvas initially. I was learning and doing. Try the game out, at least, before you continue with this site.

An arbitrage calculator for betting across bookmakers. It shows whether you win, lose, or break even from two opposite dependent events based on odds.

General Advice: Gambling is rough. Arbitrage might sound smart but still risky. If you play, be canny, not desperate. Everything’s a low-key gamble anyway.

A single-paged responsive site I put up when I was starting out with web development. I didn't use a library or framework for building; its built with vanilla HTML, CSS, and JavaScript.

I found it hard to write Hex color codes. Maybe I was too lazy to learn how to, but since I've got the abilities, why not build a color picker?

See?...That's how workarounds work.

An algorithm that checks 2 sentences to tell their similarity. It's a basic static algorithm, with different modes.

Amongst early projects on Natural Language Processing (NLP).