Posts  / #POST-216330
REDDIT

My take on a modern computer assisted codebook.

-
Jan 8, 2026 · 21:15

I've always been interested in cryptography, coding, computer science. Unfortunately I just have never had the time to learn how to code and went a completely different career route.

Full disclosure, I had lots of help from OPUS 4.5 creating this modernized codebook program.

I would like to share to hopefully get some constructive criticism so I can improve this modernized codebook.

It is executable with a stock distribution of python. No other dependencies should be needed. To my surprise, it is completely functional and does exactly what I wanted it to do!

Historical codebooks come with quite a few setbacks.

* Hard to create the initial code book
* Hard to generate new keys
* Hard to exchange keys
* Hard to query (manual)
* Weakness to frequency analysis
* Static, hard to modify
* No key rotation
* Big bulky book

My modern codebook sets out to solve all these problems

* data mine all the unique entries, common phrases, names, places from a LLM.

* Easily and securely shuffle the database creating a completely unique and one of a kind key. Fisher-Yates shuffle with secrets.randbelow() uses OS entropy and is not derived from a seed. Each message ID reaches the full 2m! Permutation equally.
Importing and exporting the shuffle maps is easy for sharing keys.

* SQLite database has high capacity to call and query the database to find the associated ID instantaneously.

* Including duplicates and randomly selecting them mitigates any frequency analysis.

* Easily add custom entries to curate your own database.

* Ability to generate key schedule files for long periods of times.

* Portable! Runs on termux.

What really sets 'ed' apart from other historical codebooks is the vast included database of large common phrases. This basically flattens any sort of potential frequency analysis to zero.

[Compose](https://postimg.cc/7Gg419sC) functiom really makes it easy to craft messages while also understanding how many duplicate IDs there are in the database.

[Here](https://gitlab.com/here_forawhile/ed/) is the code repository with a included 2million+ words/phrases database.

Please see the readme doc for a full rundown on all the commands.

This is not meant to be a replacement for modern encryption, but a exercise exploring the capabilities of modern mobile computing. That said, I cannot concieve of how eve could possibly gain any useful information from the plaintext. What is my small brain missing! What's is my Blindspot?

Post image