Appearance
Introduction
What AdminVora is
AdminVora is a production-grade admin UI system — not a pile of HTML pages with a stylesheet stapled on. It consists of four independently consumable layers:
┌───────────────────────────────────────────────────────────────┐
│ 4. Page Catalog prebuilt screens (HTML) │
├───────────────────────────────────────────────────────────────┤
│ 3. Component Library components (SCSS + JS) │
├───────────────────────────────────────────────────────────────┤
│ 2. Theme Engine tokens, every preset, layout modes, RTL │
├───────────────────────────────────────────────────────────────┤
│ 1. Runtime Core theme engine, scan/init, lazy loader │
└───────────────────────────────────────────────────────────────┘What this build is not. Earlier drafts of this page described a 120+ screen catalogue, 120+ components, and Bootstrap 5.3 compiled from source underneath. The real figures are 72 and 201 — below the original ambition on screens, above it on components — and there is no Bootstrap in this build: the component layer is an independent
.av-*system, so class names here are.av-btnand.av-inputrather than.btnand.form-control. Known limitations lists every gap.
You can adopt layer 1 + 2 only and build your own components. You can adopt all four and ship an application in a week. Nothing forces an all-or-nothing decision.
Who it is for
| Audience | What they get |
|---|---|
| Product teams | A complete, accessible, themeable UI kit so they never build a date-picker again |
| Agencies | A white-label shell that reskins to a client brand by editing ~20 token values |
| Backend developers | Server-rendered-friendly markup with zero mandatory build step |
| Frontend engineers | Real SCSS source, ES modules, TypeScript declarations, tree-shaking |
| Design teams | A token file that round-trips with Figma Variables |
What makes it different
Most admin templates on the market share the same five failures. AdminVora fixes each one deliberately.
1. They are a stylesheet, not a system
Typical templates hard-code hex values across thousands of CSS rules. Rebranding means find-and-replace across a 2 MB file.
AdminVora: every visual decision resolves through a three-tier token pipeline (primitive → semantic → component). Changing the brand hue is one token. See Design Tokens.
2. They ship jQuery in 2026
Many "modern" templates still require jQuery for their data table, their select, and their slider — adding ~90 KB before a single row renders.
AdminVora: the core runtime is zero-dependency vanilla ES2022. There is no jQuery anywhere in the product, and no adapter that would introduce it.
3. They load 40 CDN requests
CDN dependencies break behind corporate firewalls, violate strict CSP policies, fail GDPR review (Google Fonts), and rot when a version is unpublished.
AdminVora: nothing is loaded from a third-party origin. Fonts, icons, chart libraries, and polyfills are vendored into the repository by a reproducible sync tool. See Asset Vendoring.
4. They load everything on every page
A login screen should not download a charting library.
AdminVora: heavy vendors are separate files, imported only when a page has an element that needs one. A login page never downloads a charting library.
5. They are inaccessible
Custom dropdowns with no keyboard support, 2.8:1 contrast "muted" text, focus outlines removed globally.
AdminVora: an axe-core sweep at WCAG 2.1 AA over every page in both colour modes is a release gate, held to a ratchet baseline that fails on a regression and on an unrecorded improvement. Contrast is checked separately across every preset. What that does and does not cover — the 2.2 delta, screen-reader testing, the manual checklist — is set out in Accessibility.
Explicit non-goals
Being clear about what AdminVora will not do is as important as what it will.
- It is not a JavaScript framework. There is no virtual DOM, no reactivity in templates, no router. It enhances markup that your backend or framework already produced.
- It is not a CSS framework, and it does not sit on one. There is no Bootstrap, Tailwind or utility layer to learn: the component layer is an independent
.av-*system, and a page is built from named blocks (.av-card,.av-table,.av-btn) rather than from utility classes composed in the markup. If you were expecting.d-flex,.rowand.btn, they are not here. - It does not own your data layer. No fetch wrapper, no auth SDK, no ORM. Components take data; they do not fetch it. (Optional adapters exist, but nothing depends on them.)
- It is not "one template for every industry". Ten dashboards ship, but the system is optimised for teams who will build their own screens, not for people who want to ship the demo.
- It does not support IE11 or any browser without CSS Custom Properties and
:has()fallbacks. See Browser Support.
Distribution formats
| Format | Use when | Entry point |
|---|---|---|
| Source | You want to change tokens, compile SCSS, tree-shake | src/scss/adminvora.scss, src/js/adminvora.js |
| Dist | You want a drop-in <link> and <script> | dist/css/adminvora.min.css, dist/js/adminvora.js |
| HTML pages | You want the demo screens as a starting point | html/**/*.html |
Version and support policy
AdminVora follows Semantic Versioning. The av-* class names, data-av-* attributes, --av-* custom properties, and the AdminVora.* JS API form the public contract. Anything prefixed _av- or documented as internal may change in any release.
Read next: Known Limitations.