Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/iamk77/skill/colornpx skills add IamK77/Skill --skill colorgit clone --depth 1 https://github.com/IamK77/SkillWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00220 | $0.05213 |
| Opus 5 | $0.00110 | $0.02606 |
| Sonnet 5 | $0.00044 | $0.01043 |
| Haiku 4.5 | $0.00022 | $0.00521 |
Grade A, and why
color scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 170 lines — stays where its author put it; the contents beside it link to each section on GitHub.
color
!checklist init ${CLAUDE_SKILL_DIR} --force
An atelier (a craftsman's studio) is named for the discipline of the hand — careful, deliberate, derived. color is the color & light lens of the atelier suite: where the other skills decide what is true and what is composed, color decides what every hue, tone, and value actually is — and makes that a system, not a scatter of hand-picked hexes. Its product is a tokenized color system: a perceptual neutral ramp with a chosen temperature, semantic role tokens mapped to that ramp, a correctly engineered dark theme, contrast verified at design time, and a data-viz scale engineered for perceptual uniformity and colorblind safety — each committed before components consume it. It runs across gated stages and will not advance past a GATE until the checklist tool clears it — order enforced, substance yours.
The governing fact: color coherence is not a matter of taste applied per component — it is a matter of deriving every value from a small, decided system. The same blue typed as #2563eb in one component and #2f6ae0 in another, a neutral gray that is just #888888, a dark mode that runs filter: invert(), a data viz that uses a stock rainbow — each is a hand-picked value, and their sum is the perceptual incoherence the eye reads as "off" without being able to name it. The craft is the same subtraction the suite preaches everywhere: define a small system (one perceptual space, one neutral temperature, a handful of semantic roles, one contrast target), then derive every concrete value from it. Do that and two things happen — the palette becomes coherent (a thousand component colors agree because they came from ten tokens), and it becomes evolvable (dark mode, re-theme, or rebrand by remapping the alias layer, not finding-and-replacing scattered hex).
This is where the agent era bites:
- The agent picks plausible-but-incoherent values and feels no wrongness. It will emit
oklch(0.62 0.18 250)as a primary blue andoklch(0.60 0.17 248)for success green — close enough to look intentional, wrong enough to read as undesigned. It has no nervous system to flinch at a secondary text color that fails WCAG. The values must come from a decided system, not per-call taste. - The agent reaches for the cheap trick. Dark mode by
filter: invert(), a rainbow palette for a data series,color: grayfor disabled,border-color: #cccper component. Each is the move that looks like the craft and isn't. - The agent maintains two sources of truth without noticing. It defines
--color-primary: oklch(0.62 0.18 250)in CSS and then hardcodescolor: '#2563eb'in a React inline style — two palettes, one in each language, that will silently drift. The design system is itself state; it gets one source of truth (seeatelier:systems).
Speak the user's language. The decisions are the user's — which neutral temperature (warm / cool / brand-tinted), which accent hue, what the contrast targets are, whether the surface skews editorial or product. Read their fluency and gloss a term on first use (perceptual color / oklch, neutral ramp, semantic token, alias layer, contrast ratio / APCA Lc, colorblind-safe / CVD). A token system the user can't read is a palette imposed, not shared.
What ships with it
12 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- .checklist.yml 10 KB
- LICENSE 11 KB
- NOTICE 703 B
- probes/color-coherence.mjs 12 KB runs code
- probes/README.md 3.7 KB
- probes/test.mjs 8.1 KB runs code
- probes/TESTING.md 2.7 KB
- references/color-spaces-and-neutrals.md 9.1 KB
- references/contrast-and-bw.md 10 KB
- references/dark-theme-engineering.md 13 KB
- references/data-viz-color.md 14 KB
- references/oklch-palette-recipes.md 11 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 170 lines · 220 tokens per session scan A d88c2faf7666
color is a skill published in the GitHub repository IamK77/Skill (2 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 220 tokens to every session and 5,213 once invoked, about $0.0011 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
fixing-motion-performance
Audit and fix animation performance issues including layout thrashing, compositor properties, scroll-linked motion, and blur effects. Use when animations stutter, transitions jank, or reviewing CSS/JS animation performance.
web-design-engineer
Build or redesign polished browser-rendered visual artifacts with HTML/CSS/JavaScript/React: pages, dashboards, prototypes, slide decks, animations, UI mockups, and data visualizations. Use for visual front-end creation, design-system exploration, design critique, or explicit browser acceptance / QA of a web artifact.…
skillshare-ui-website-style
Skillshare frontend design system for the React dashboard (ui/) and Docusaurus website (website/). Use this skill whenever you: build or modify a dashboard page or component in ui/src/, style or layout website pages or custom CSS in website/, create new React components for the dashboard, add pages to the dashboard…
baseline-ui
Quickly deslop UI code by fixing spacing, hierarchy, typography, and small layout issues. Use when the interface needs a fast cleanup or polish pass.
design-artifact
Design principles and creative direction for building HTML artifacts — pages, reports, plans, landing pages, demos, decks, and small tools. Use when creating or restyling any visual HTML deliverable and deciding its palette, type pairing, layout, theming, or overall register, or when the output must not look…
frontend-design
创建具有高设计品质、可交付生产的前端界面。当用户要求构建 Web 组件、页面、海报或应用(如官网、落地页、仪表盘、React/Vue 组件、HTML/CSS 布局,或对任意 Web UI 进行样式/美化)时使用本技能。产出有创意、打磨到位且避免「AI 通用审美」的代码与界面设计。.