Brand kit
Put the person's colors, fonts, and usage rules on one page. Then prove the kit works by restyling one real artifact. Use when they say "make my stuff look consistent", "brand guidelines", or their site, slides, and posts all look unrelated.From its SKILL.md
npx -y skills add protosphinx/sphinxstack --skill brand-kitAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
2.9 KB, 632 tokens by cl100k_base, as published. Nobody here has run it
brand-kit
Write down the visual decisions for someone's real project so they stop re-deciding them. The output is one page — a brand kit — plus one existing artifact restyled to match it, which is the proof the kit works. This assumes some decisions already exist or get made now; if they have nothing at all, run pick-fonts-and-colors first and come back.
Collect what exists
Look at everything they currently have: site, resume, slides, avatars, posts. List every font and color actually in use (pull hex values from the real files, do not guess). Show them the list — seeing eleven grays and five fonts is usually the moment they understand why a kit exists.
Decide and cut
Work down to a small fixed set, their pick at each step with your one-line trade-off on the options:
- Two fonts: one for headings, one for body. Free sources only (Google Fonts). Name the exact weights allowed — usually two per font.
- Colors: one background, one text color, one accent. Optionally one neutral gray for secondary text. Exact hex values. Check the text/background pair for contrast (aim for WCAG AA) and say so in the kit.
- Logo usage if a logo exists: minimum size, clear space, the black and white variants, and the one background it never goes on.
- Voice line: one sentence on how written things sound, in their words.
Everything not on the list is banned by default. That is the point of the kit — the next decision is already made.
Write the page
One file, brand.md (or brand.html if they want it visual), in
the project repo. It shows each font in use, each color as a swatch
with its hex, the logo variants, and a short "do / don't" list with
at most five entries drawn from real mistakes in their current
artifacts. Keep it under a page. A kit nobody rereads in thirty
seconds is a kit nobody uses.
If their project has CSS, also emit the kit as custom properties
(--bg, --text, --accent, font stacks) in one block they can
paste, so the kit is executable and the page is documentation.
Prove it
Pick the artifact of theirs that currently strays furthest from the kit and restyle it together, using only what the kit allows. Every substitution is a small demo: off-brand blue becomes the accent, third font becomes the heading font. Ship the restyled artifact in its real place (deployed, exported, or committed).
Done
- The kit page exists in the repo and records fonts, colors, logo rules, and voice
- One real artifact matches the kit and passes the stated contrast check
- The person can answer a common design question from the kit without reopening a color or font picker
What ships with it: 1 file
208 B alongside SKILL.md
agents/
- openai.yaml208 B
Gives 0 of the 12 instructions most marketing audience skills give in 632 tokens
Counted across 690 of the 894 authors here whose files we hold, read 2026-08-07
- Apply Poppins font to headingsin 41 of 690, across 6 files
- Apply Lora font to body textin 41 of 690, across 6 files
- Use Arial fallback for headingsin 39 of 690, across 4 files
- Use Georgia fallback for body textin 39 of 690, across 4 files
- Maintain text hierarchy and formattingin 39 of 690, across 4 files
- Use accent colors for non-text shapesin 38 of 690, across 3 files
- Use RGB values for precise color matchingin 38 of 690, across 3 files
- Use brand colors for primary text and backgroundsin 36 of 690, across 1 file
- Read product marketing context file before asking questions, starting, or auditingin 35 of 690, across 23 files
- Use active voice instead of passive voicein 26 of 690, across 10 files
- Implement or generate appropriate JSON-LD structured datain 24 of 690, across 17 files
- Prioritize clarity over clevernessin 22 of 690, across 8 files
Said here and by no other author read
- list fonts and colors from existing files
- do not guess hex values
- use free font sources only
- limit to three core colors
- check text/background pair for WCAG AA contrast
- state the contrast check result in the kit
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.