Adopt DESIGN.md format for visual identity documentation

[EthanThatOneKid]

proposal

Adopt the DESIGN.md format to document the acmcsuf.com visual identity in a machine-readable way.

why

DESIGN.md is a format spec that combines YAML front matter (design tokens) with markdown prose (design rationale). It lets AI coding agents understand a design system persistently and structurally — rather than inferring it from prose or screenshots.

For example, agents could:

• Generate new UI components that match the brand automatically
• Validate new designs against the brand tokens
• Diff brand changes over time
• Export tokens to Tailwind, Figma, or other tooling

what to add

A DESIGN.md file in the repo root covering:

1. Colors — brand palette (primary, secondary, accent, neutral, etc.)
2. Typography — font families, sizes, weights, line heights
3. Spacing & rounded — scale tokens
4. Components — button styles, form elements, etc.
5. Prose sections — Overview, Do's and Don'ts, etc.

relevant links

• Spec: https://github.com/google-labs-code/design.md
• npm CLI: @google/design.md
• Export to Tailwind or W3C DTCG format supported

next steps

1. Audit the existing brand assets (CSS, Figma, docs) for token values
2. Draft a DESIGN.md file
3. Validate with npx @google/design.md lint DESIGN.md
4. Optionally wire it into the CI to catch brand regressions

Would love feedback from the team on whether this fits the current workflow!

[GaballaGit]

I am going to throw in my 2 cents. I think I am actually opposed to this proposition.

I feel that the best part about participating in ACM@CSUFs open source projects is how you learn and get mentored by more experienced contributors, typically the board officers. Now, when we introduce DESIGN.md we are opening up the doors for better agentic ai flow. Which could be a good or bad thing depending on who the user is. However, I think this is not the direction we should take as this may cause future contributors to just take shortcuts on their issues with something like "claude write me this whole ahh page" or something, which in turn leads to them not learn as much because 1. they aren't writing the code, and 2. they aren't talking to more experienced people.

I think rejecting a more convenient agentic ai workflow will help push newer contributors to at least learn a bit even if they decide to use an agentic ai regardless. For example in this case with no DESIGN.md, contributors would need to study the code base or ask someone for help with the styling. With DESIGN.md I feel that this step would be eliminated and contributors would have miss a chance to develop their skills related to contributing to open source projects as a whole.

That how I see it, but maybe I hold the minority view here.

[GaballaGit]

cc @acmcsufoss/board