Understand

Color debt linter

Find the colours in your CSS that are the same colour under different names.

What color debt is

25%

colour debt

3 of 12 colours could be deleted without anyone seeing a difference.

Distinct colours

Declarations

Files scanned

The whole palette, most-used first

ΔE < 2.3 — the same colour in practice. Keep one, replace the rest.

Keep

Replace #0F172B

−1

#0f172b·1 use·src/app/globals.css:1

Keep

Replace #314158

−1

#314158·1 use·src/app/globals.css:1

Keep

Replace #90A1B9

−1

#90a1b9·1 use·src/app/globals.css:1

Take the report with you

Untimestamped, so the same files always produce a byte-identical report — commit it, diff it, paste it where the decision is actually being argued.

## Colour debt

**3 of 12 colours are removable** (25%), across 20 declarations in 3 files.

### Duplicates (ΔE < 2.3 — the same colour in practice)

| Keep | Merge | Uses | First seen | ΔE spread |
| --- | --- | ---: | --- | ---: |
| `#0F172A` | `#0F172B` | 1 | `src/app/globals.css:1` | 0.54 |
| `#334155` | `#314158` | 1 | `src/app/globals.css:1` | 1.19 |
| `#94A3B8` | `#90A1B9` | 1 | `src/app/globals.css:1` | 1.27 |

### Near-misses (ΔE < 5 — visible, but not reliably)

| A | B | ΔE | First seen |
| --- | --- | ---: | --- |
| `#2563EB` | `#3070F0` | 4.51 | `src/styles/tokens.legacy.css:1` |

Thresholds: ΔE < 2.3 is the CIEDE2000 just-noticeable difference; ΔE < 5 is visible under comparison but not across displays. Survivors are always colours already in the codebase, never averages, so the diff stays reviewable.

All tools

Palette DoctorWhat's wrong, in plain words Contrast CheckerWCAG + color-blind checks Site InspectWebsite → colors, fonts, assets

Take the report with you

Keep going