Tools / DESIGN.md check
DESIGN.md check Free
Scores a DESIGN.md (the brand and design context file models read before building): missing sections, no hex values, text pairs that fail contrast, vague adjectives, no never-list, too long.
MCP tool name: lint_design_md. REST: POST /api/design-md-lint. Free: 30 runs a day per IP, 20,000 characters each. Pro: no cap.
Ask for it in plain words
Score a DESIGN.md (the brand and design context file models read before building): missing sections, no hex values, text and background pairs that fail contrast, too many typefaces, adjectives a model reads as the generic look, no never-list, no sizes, too long or too thin.
Inputs
| Name | Type | Meaning |
|---|---|---|
text * | string | The DESIGN.md contents |
* required
Example
curl -X POST https://designers.allthepossibles.com/api/design-md-lint \
-H "content-type: application/json" \
-d '{"text":"# Acme design\n\n## Overview\nA clean, modern, friendly app for teams.\n\n## Colour\n- Text: #9a9a9a\n- Background: #ffffff\n- Accent: #6366f1\n\n## Typography\nInter for everything."}'
Example output
The report field, for the input above (structured fields come back alongside it):
DESIGN.md score 4/100: 21 words, 3 of 9 expected sections, 3 colours, 1 text and background pairs checked
Sections: ok purpose and audience; missing principles or voice; ok colour; ok typography; missing spacing and layout; missing components; missing imagery and motion; missing accessibility; missing do and do not
HIGH Text #9a9a9a on Background #ffffff is 2.81:1; text needs 4.5:1
HIGH no "never" list; the fastest way to stop generic output is naming the defaults you reject (gradients, eyebrow labels, three identical cards)
MEDIUM no principles or voice section; a model has to invent it, and it will pick the default
MEDIUM no spacing and layout section; a model has to invent it, and it will pick the default
MEDIUM no components section; a model has to invent it, and it will pick the default
MEDIUM no imagery and motion section; a model has to invent it, and it will pick the default
MEDIUM no accessibility section; a model has to invent it, and it will pick the default
MEDIUM no do and do not section; a model has to invent it, and it will pick the default
MEDIUM adjectives a model reads as the generic look (clean, modern, friendly); say what you mean with a value, a reference or an example
MEDIUM only 21 words; too thin to constrain a model
MEDIUM no sizes or spacing values; give the type scale and the spacing unit
LOW no references or assets linked; one real example beats a paragraph of adjectives
A DESIGN.md is read by a model before it builds, so it is judged by whether it changes the output: concrete values, named rejects and real examples.
What it looks like
Limits of this tool
Deterministic: the same input always gives the same answer. It does not call a language model.