From First Release to Better Release: Confronting My Tool’s Flaws

In December 2025, I built and published a username reconnaissance tool on GitHub and PyPI. Along with that, I also shared the tragedy-cum-comedy story of how the tool was developed and finally published.



For the past month, I was away from tool development and core cybersecurity studies due to some personal reasons. But this week, I decided to revisit the project and test it again.

And that’s when reality hit 😅.

During testing, I was able to identify numerous problems in the tool. Obviously, these issues were not visible at the beginning  because at that time, everything felt like a eureka moment.

What Should an Engineer Do When Problems Appear?

The obvious answer is:
fix them.

But engineering is not about solving all problems at once.

The real philosophy is:

"Solve one problem properly, then move to the next.
Eventually, you conquer Everest 😎." 

Problem #1: Platform Selection

The first issue I wanted to solve was lack of flexibility.

Earlier, the tool checked usernames on all platforms by default. I wanted to give users control.

So I planned two options:

  1. Check all platforms

  2. Check a specific list of platforms

This was easy to implement, and I did it.

But engineers rarely stop at “easy”.

Improving the Idea

I shared this idea with DeepSeek to see if there was a better approach.
The suggestion I got was actually smarter:

When a user chooses specific platforms, the tool should display the full list of available platforms and let the user select from it.

This feature turned out to be extremely useful. Now users know exactly:

  • which platforms are supported

  • and where their username is being checked

Moreover to this, now users can actually save the result in a file with a single click.

That problem was officially solved ✅.

Publishing the Update (GitHub & PyPI)

Next step: publishing the new version.

GitHub

I already had a repository with the same tool name. I asked ChatGPT how to handle updates.

Answer was simple:

Just update the relevant files and create a new release. No need to change anything else.

GitHub release done 😍.

PyPI: Comedy Continues (Again)

Publishing on PyPI was… interesting.

This time it wasn’t a tragedy — more like a controlled comedy 🎭.

I remembered some steps, did whatever I could, and then took help from ChatGPT. Within 10 minutes, a new version was uploaded to PyPI.

Of course, no deployment is complete without errors.

The Usual Bug (Again 😅)

While testing, I hit the familiar error:

No module named platform

I already knew the fix this time, so it was resolved in 3 minutes.

But this was not the main issue.

The Real Problem: Old Versions Getting Installed

No matter what I did, pip install userrecon kept installing the older version.

This confused me. I asked ChatGPT again and learned something new.

Important Learning

pip caches old versions by default.

Why?

  • Wheel files are cached

  • Windows sometimes doesn’t refresh executable scripts

The Fix

Two simple steps solved everything:

pip uninstall userrecon -y
pip install --no-cache-dir userrecon

Now the latest version was installed properly.

Final Test: Another Machine

If you remember the conclusion of my previous blog:

“If it works on your machine, it doesn’t mean it will work everywhere.”

So I decided to test it on someone else’s laptop.

As soon as I returned to my apartment from the library, I borrowed my friend’s laptop and ran:

pip install userrecon

And… Bingo 🎉
The latest version v1.1.0 was installed successfully 🏆.

Final Thoughts

This experience reminded me why revisiting your own work is important.

Initial success feels great — but real engineering begins after that, when you:

  • notice flaws

  • improve usability

  • fix deployment issues

  • and test beyond your own system

I’ll continue improving the tool step by step.
One problem at a time.


Comments

Post a Comment

Popular posts from this blog

Obfuscation and Deobfuscation

I was on Hack The Box Academy, happily reading along when the first word of the module popped up: understanding . Good. Next word: code . Still good. Then the third word appeared: deobfuscation . And I just stared at it like, “How many alphabets did they use to make this word?” I tried pronouncing it five or six times. For a moment, I genuinely believed Shashi Tharoor writes simpler words 😂. Eventually, yes, I managed to say it without hurting my tongue. Anyway, let’s move to the topic. Obfuscation  As the name suggest , it is process of converting simple things ( notes , words , sentences ) into complex form 😂.  More precisely , it is the process of taking a simple , clear sentence and writing it in a complex cipher, using synonyms from a dead language, and adding alot of irrelevant words. The meaning remains preserved, but it is extremely hard to understand ( same as the name of topic obfuscation) .  Defuscation It is simply opposite of obfuscation. It is the proc...
Read more

OSINT Basics: Introduction, Scope, and Ethical Boundaries (Part 1)

Image
Disclaimer: All the tools and methods discussed in this blog are part of what I am learning in a safe and legal environment . This content is strictly for educational purposes only and is not intended to harm any individual or organization. Even though we all know that theoretical knowledge alone is not enough, most of my learning so far has been on the theory side. From now on, I’ll try to balance both theory and practical understanding — slowly and responsibly. That brings us to today’s topic: OSINT . What is OSINT? OSINT stands for Open Source Intelligence . As the name suggests, it is about gathering information from publicly available sources such as: the internet social media news articles public records OSINT has nothing to do with hacking . It focuses on collecting and analyzing information that is already exposed — often unintentionally. Why OSINT matters in cybersecurity OSINT helps us understand the human attack surface , which is often the weakest link in sec...
Read more

Cybersecurity Devices and Technologies Part 1

There is no single device or technology that can solve all the security needs of an organization on its own. A secure network is built by choosing the right tools for the right purpose. This blog covers some of the basic devices and technologies used in the field of network security. Security Appliances Security appliances can be physical devices like routers, or software tools running on those devices. They generally fall into six categories: Routers : Routers primarily connect different network segments, but they also offer basic traffic-filtering features. This helps decide which devices can connect and which networks they’re allowed to communicate with. Firewalls : Firewalls are the guardians of a network. They inspect traffic more deeply and can detect suspicious or harmful behaviour. Firewalls enforce security policies on every packet that passes through them. Intrusion Prevention System (IPS) : An IPS uses traffic signatures to detect and block malicious activities i...
Read more