CyberToolkit

CyberToolkit started as a hands-on way to work with sockets, hashing, and the everyday building blocks of network security work. I kept it standard-library only, so every piece, from socket handling to thread pooling, is something I had to understand rather than import.

• Port scannerConcurrent TCP scan with ranges
• SHA-256 file hashChunked, so large files work
• Password generatorBuilt on secrets, not random
• DNS lookupAll IPs for a domain
• Banner grabberPrints the service banner
• Python 3.8+No dependencies

Output from the example in the project README. scanme.nmap.org is a host Nmap provides specifically for testing.

• Threaded scanning. A sequential scan of 1,000 ports at a 0.5s timeout could take minutes when ports are filtered. A ThreadPoolExecutor overlaps the waiting, so the scan is bounded by the slowest port rather than the sum of all of them.
• secrets, not random. Python's random is a pseudo-random generator and unsafe for anything security-related.
• Chunked hashing. Files are read 64 KB at a time, so hashing doesn't depend on a file fitting in memory.
• Specific exception handling. Each tool catches the errors it actually expects and reports them distinctly, instead of a bare except that hides problems and swallows Ctrl+C.
• Hostnames resolved once. A bad hostname fails immediately instead of erroring on every port.

A pytest suite covers the port-specification parser: ranges, lists, de-duplication, and invalid input. The network functions are verified manually against a local listener.

Public on GitHub under the MIT license. A separate lab copy, cybertoolkit-lab, is where I experiment with AI coding agents without touching the main repository.

Possible extensions

• UDP scanning
• Export results to JSON or CSV
• Concurrent banner grabbing across multiple ports
• Reverse DNS lookups