I’m building a Prime Number Tutor on Manjaro Linux alongside an AA Trainer dashboard that checks whether the project is ready to publish. The tutor explains whether a number is prime or composite in plain language, while the dashboard helps me track documentation, privacy checks, screenshots, package files, release status, and other details that make software easier to trust and maintain.
I’m preparing the project for GitHub, a genuine educational dataset on Hugging Face, and Arch/Manjaro-compatible packaging, including Snap distribution. I deliberately block automatic installation of untrusted AUR or other user-repository code, while still allowing my own package files to be prepared for publication. The project uses Apache-2.0 so developers and future software or AI tools can lawfully reuse the code, although making code public does not guarantee that any AI company will train on it.
A growing part of the project is automation: privacy checks, documentation checks, release verification, checksums, public metadata, packaging, and testing before software moves from an early release toward a stable one. The larger purpose is to learn how software can be built carefully from scratch and then made understandable and useful to other people; I wrote more about that at https://j03.page/2026/09/01/building-software-from-scratch/. I’ll keep documenting what works, what breaks, and what I learn along the way at j03.page.
Leave a comment