CSS
Collect styles from pages
You provide a URL. Milten parses available CSS sources on the page and extracts values that can be compared by frequency: colors, font-family, font-size, sizes, spacing, radii, shadows, and custom properties.
CSS Tokens scanner
Milten builds a frequency summary of repeated design values in CSS: colors, fonts, sizes, spacing, radii, shadows, and custom properties.
How the scanner works
From CSS sources to a frequency summary of values
Frequency
Group repeated values
The report counts how often identical colors, fonts, sizes, spacing, border-radius, box-shadow, and CSS variables appear. The more often a value repeats, the more it looks like a system token.
Summary
Show concentration and spread
The summary helps you quickly see where values repeat consistently and where the interface has accumulated a wide spread of colors, sizes, fonts, or shadows.
Reading
Help choose what to review first
The report does not replace a designer's decision, but it gives a practical order: start with frequently repeated values, and leave rare or empty groups for later.
What is in the report
Seven value groups counted by the worker
Colors
Repeated colors
The scanner shows color values and their usage counts so you can separate system colors from accidental local shades.
valueuses
Fonts
Fonts and text sizes
The report includes detected font-family values, how often they appear, and repeated font-size values.
font-familyfont-size
Spacing / size
Sizes, margin, padding, width, and height
The summary highlights frequently used size and spacing values. A large spread here usually points to weak grid consistency.
pxrem
Radius / shadows
Radii and shadows
Separate groups show repeated border-radius and box-shadow values, including cases where one value appears more often than the others.
radiusshadow
CSS variables
Detected custom properties
The scanner shows CSS custom properties and their usage counts so you can see which variables already work like tokens.
--varuses
FAQ
Frequently asked questions about the CSS Tokens scanner
Short answers about which values enter the report and how to read the frequency summary.
How is the CSS Tokens scanner different from exporting CSS variables?
It looks wider: beyond custom properties, the report shows repeated colors, fonts, font-size values, sizes, spacing, border-radius, and box-shadow. Usage frequency matters for every value.
Can I analyze a site without repository access?
Yes. The first pass can run against public pages and linked styles. If part of the CSS is behind authentication or does not load on the page, it will not enter the summary.
Does the report show token conflicts?
The current worker does not infer semantic conflicts or suggest token names. It shows repetition and value spread, while tokenization decisions stay with the team.
Is the scanner useful if tokens already exist?
Yes. The report shows which CSS variables were found and how often they are used, plus where repeated raw values still exist nearby: colors, sizes, spacing, or shadows.
How should I read empty or rare values?
Empty groups and one-off values are usually less important. The report primarily helps find repeated values that make tokenization easier to start.
Related services
What to check next to this service
Adjacent checks help reveal nearby causes across speed, code, accessibility, SEO, and page state after user scenarios.
CSS audit
Stylesheet size, unused selectors, duplicates, and specificity explosion. See what can be removed safely.
OpenHTML validation
Check page structure, validity, semantics, and required meta tags.
OpenAccessibility (a11y)
WCAG audit: contrast, alt text, landmarks, aria attributes, and keyboard navigation. Includes fixes and examples.
OpenLoad speed
Check Web Vitals and key speed metrics. The scanner runs two passes: mobile and desktop, then returns a prioritized fix plan.
OpenStart with repeated CSS values
Paste a URL and get a summary of colors, fonts, sizes, spacing, radii, shadows, and CSS variables that actually appear on the site's pages.
Get token summary