Since diving into CTFs more seriously, I found myself stuck in the same loop:
🔍 Run Nmap
📄 Read the results
🤖 Ask GPT for insights
One day I thought:
"Why not automate this?"
What if I could create a tool that runs Nmap in the background and feeds the output directly into an AI agent?
That was the spark. I started chatting with GPT to figure out how to approach it. My goal wasn’t to reinvent Nmap — I wanted to build on top of it, keeping all its power intact.
I chose Python to build the CLI, keeping things simple. The flow looked like this:
Run Nmap with some default parameters
Parse the results and organize them into a clean table
Send a prompt with context to an LLM (I used Gemini because it's free 😄)
Make an API call to Vulners to look up known vulnerabilities
Add a final touch: some ASCII art for fun 🎨
And that was it — I ended up with a tool that pulls recon data, enriches it with external sources, and asks an AI to help me interpret it.
It was a super fun project, and I learned a lot by building it from scratch.
🚀 Curious? Check it out at the link below!
Top comments (0)