Run2 anthropic brand tokens
[COLM'26] SkillLearnBench is the first benchmark for evaluating continual learning methods that automatically generate agent skills.
npx -y skills add cxcscmu/SkillLearnBench --skill run2_anthropic-brand-tokensAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
What its author says it does
Copied from the file, not written here
Precise Anthropic brand color tokens and typography guidelines for design artifacts, with matplotlib-specific notes.
SKILL.md
2.8 KB, 706 tokens by cl100k_base, as published. Nobody here has run it
Anthropic Brand Tokens (Refined)
Color Palette
| Token Name | Design Role | HEX |
|---|---|---|
| Identity Light | Page / poster background (never pure white) | #FAF9F7 |
| Corporate Dark | Outer casing, headings, dark UI elements | #1A1816 |
| Primary Brand Accent | Connectors, CTAs, interaction highlights | #CF6E35 |
| Secondary Brand Accent | Thermal management, supporting hardware | #8C9FA0 |
| Tertiary Brand Accent | PCB substrate, circuit-board elements | #7A8B72 |
| Muted Mid Gray | Annotation leader lines, dividers | #9B9B9B |
Supporting Colors (derived, not brand tokens)
| Name | Usage | HEX |
|---|---|---|
| Battery Warm Neutral | Battery pack layer | #C8BFB4 |
| Metadata Bar | Footer/header subtle bar | #EDEAE5 |
Color Rationale
- Identity Light (#FAF9F7): warm off-white; the subtle sepia undertone prevents sterility
- Corporate Dark (#1A1816): very dark warm charcoal; never cold pure black
- Primary (#CF6E35): Anthropic's terracotta/coral signature — used sparingly for maximum impact
- Secondary (#8C9FA0): muted teal-gray — a cool industrial complement for thermal hardware
- Tertiary (#7A8B72): muted sage/olive — organic feel suits PCB green substrate
- Mid Gray (#9B9B9B): exactly mid-brightness neutral for structural lines
Typography
matplotlib Font Guidance
- System font for matplotlib on Linux:
DejaVu Sans(always available, sans-serif) - Preferred when installed:
Inter,Helvetica Neue, orArial - Check available fonts:
from matplotlib import font_manager; [f.name for f in font_manager.fontManager.ttflist]
Letter-spacing in matplotlib
- IMPORTANT:
letter_spacingis NOT a valid matplotlib Text property — it will raise AttributeError - Simulate letter-spacing by inserting spaces between characters:
"N O V A"instead of"NOVA" - For large display titles, spaces of 1–2 characters work well
Text Hierarchy
| Level | Size | Weight | Color |
|---|---|---|---|
| Display title (NOVA) | 52–60pt | bold | Identity Light on dark bg |
| Subtitle | 9–10pt | light | Muted Mid Gray |
| Layer label | 9pt | regular | Corporate Dark |
| Annotation text | 8pt | light | Corporate Dark |
| Metadata | 7.5pt | regular | Muted Mid Gray |
Brand Constraints
- Background MUST be Identity Light (#FAF9F7), never pure #FFFFFF
- No neon colors, no high-saturation gradients
- Primary accent appears only for focal interaction points (not for backgrounds)
- Low saturation overall — muted, professional, warm-neutral aesthetic
- Dark bars (title/footer) use Corporate Dark, NOT pure black
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.