How this is put together
UX Atlas is a single content model rendered through a small set of templates. That is what lets it grow without the reading experience drifting, and what makes the relationships between entries reliable rather than decorative.
One schema, many templates
Every entry, from a law to a tool listing, uses the same shape: an identity, a summary, a canonical definition, an evidence rating, a body built from typed content blocks, examples, tags, relationships and sources. Specialised types add fields on top of that rather than replacing it.
The consequence is that adding content never means adding page code. There are exactly two resource routes in the application: a section index and a resource page, with the resource page dispatching to a handful of specialised templates for tools, glossary terms, case studies, checklists and templates.
Relationships are resolved, not written twice
Entries declare relationships as identifiers. At build time the registry symmetrises them, so a link declared in one direction works in both; drops any that do not resolve, so a broken link never renders; and back-fills tag-based neighbours where an entry is thinly connected.
That is why no page here is a dead end. It is a property of the data model rather than something an author has to remember on every entry.
Evidence is labelled, not implied
Design writing routinely presents a satirical essay and a replicated experiment as though they carry equal weight. Every entry here carries one of four ratings, and the body says where the popular version of an idea overstates its source.
- Established research
- Originates in published research and has been replicated. The effect is real. Its size in your product still needs measuring.
- Mixed evidence
- Real research exists, but findings conflict, effect sizes are small, or the jump from the lab to interface design is contested.
- Expert heuristic
- A review criterion developed by practitioners. Useful and widely taught, but evaluated by judgement rather than measurement.
- Craft convention
- True because the industry converged on it. Breaking it costs familiarity, not correctness.
What this does not do
- It does not invent laws. Every named effect has a published origin, cited on the page.
- It does not fabricate research findings or attribute invented numbers to real studies.
- It does not present case studies as reports on named companies. They are composites, labelled as such, written to teach reasoning.
- It does not track you. Checklist progress, audit findings and template drafts are stored in your browser and never sent anywhere.
- It does not claim conformance on your behalf. The accessibility material is guidance, and a real audit still needs manual testing with assistive technology.
Content integrity
The registry reports problems rather than failing the build, so a content error is visible instead of silent. This page shows the current state.
- Entries
- 485
- Duplicate identifiers
- 0
- Relationships that did not resolve
- 0