sprezzature-ux-laws

🇫🇷 LISEZMOI.md · 🇬🇧 README.md

A static Laws-of-UX auditor for HTML.

The "Laws of UX" are named heuristics from psychology and human-computer-interaction research: for instance, Hick's Law says that the more choices a screen offers at once, the longer a person takes to pick one, so a navigation menu with forty links is measurably harder to use than one with seven. This tool cannot see a rendered page or watch a real person use it; instead it reads the raw HTML text and flags the mistakes that are decidable from that text alone, such as counting the links in a <nav>. That trade keeps it fast and deterministic enough to run in a pre-commit hook or a CI step, at the cost of catching only the fraction of each law's violations that a text scan can see (the full, sourced definition of each law lives in references/laws-of-ux.md).

Laws covered

Use

# Audit a file or a whole directory tree
python scripts/audit_laws_of_ux.py public/index.html
python scripts/audit_laws_of_ux.py site/ --json

# Restrict to specific laws, or auto-fix the mechanical ones
python scripts/audit_laws_of_ux.py page.html --only fitts,jakob
python scripts/audit_laws_of_ux.py page.html --fix

Findings are error or warning; the process exits non-zero when an error is present, so it drops cleanly into CI.

Honest scope

The auditor catches mechanical, source-decidable violations. It does not judge whether a screen answers the right question or whether a flow makes sense. That stays a human call. It is a fast gate, not a substitute for design review.

License

BSD-3-Clause © Warith HARCHAOUI. Part of the sprezzature toolkit.