Impeccable
bestofjs/bestofjs
A skill your agent uses when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or otherwise improve a…
This skill should be used when the user explicitly says "Lunora style", "Lunora design", "/lunora-design", or directly asks to use/apply the Lunora design system.
$ npx skills add anolilab/lunora --skill lunora-design -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install anolilab/lunora lunora-design --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/anolilab/lunora.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/lunora-design .claude/skills/lunora-design && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "lunora-design" agent skill from https://github.com/anolilab/lunora/tree/alpha/.agents/skills/lunora-design into .claude/skills/lunora-design/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "lunora-design", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/anolilab/lunora/tree/alpha/.agents/skills/lunora-designType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add anolilab/lunora --skill lunora-design -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install anolilab/lunora lunora-design --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/anolilab/lunora.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.agents/skills/lunora-design .agents/skills/lunora-design && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "lunora-design" agent skill from https://github.com/anolilab/lunora/tree/alpha/.agents/skills/lunora-design into .agents/skills/lunora-design/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "lunora-design", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add anolilab/lunora --skill lunora-design -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install anolilab/lunora lunora-design --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/anolilab/lunora.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.agents/skills/lunora-design .cursor/skills/lunora-design && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "lunora-design" agent skill from https://github.com/anolilab/lunora/tree/alpha/.agents/skills/lunora-design into .cursor/skills/lunora-design/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "lunora-design", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/anolilab/lunora.git --path .agents/skills/lunora-design--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add anolilab/lunora --skill lunora-design -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install anolilab/lunora lunora-design --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/anolilab/lunora.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.agents/skills/lunora-design .gemini/skills/lunora-design && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "lunora-design" agent skill from https://github.com/anolilab/lunora/tree/alpha/.agents/skills/lunora-design into .gemini/skills/lunora-design/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "lunora-design", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install anolilab/lunora lunora-designInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add anolilab/lunora --skill lunora-design -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/anolilab/lunora.git skills-src && mkdir -p .github/skills && cp -r skills-src/.agents/skills/lunora-design .github/skills/lunora-design && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "lunora-design" agent skill from https://github.com/anolilab/lunora/tree/alpha/.agents/skills/lunora-design into .github/skills/lunora-design/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "lunora-design", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add anolilab/lunora --skill lunora-design -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install anolilab/lunora lunora-design --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/anolilab/lunora.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.agents/skills/lunora-design .opencode/skills/lunora-design && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "lunora-design" agent skill from https://github.com/anolilab/lunora/tree/alpha/.agents/skills/lunora-design into .opencode/skills/lunora-design/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "lunora-design", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
lunora-designThis skill should be used when the user explicitly says "Lunora style", "Lunora design", "/lunora-design", or directly asks to use/apply the Lunora design system.
Lunora Design is an agent skill from anolilab/lunora. This skill should be used when the user explicitly says "Lunora style", "Lunora design", "/lunora-design", or directly asks to use/apply the Lunora design system. NEVER trigger automatically for generic UI or design tasks.
Its SKILL.md is about 4.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 6 other files, including reference files (for example `README.md`, `references/components.md` and `references/platform-mapping.md`).
It sits in Frontend & Design, covering Design systems. The repository describes itself as: Type-safe, real-time backend framework on your own Cloudflare account — Workers, Durable Objects, D1, R2, Queues. Convex-style DX, Vite-first. The licence is MIT.
5 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 8597a6b. It shows what the files ask for, not the result of running them.
Pre-approves these tools, so the agent can use them without asking each time:
ReadWriteEditGlobGrepFrom allowed-tools in the SKILL.md frontmatter.
No scripts in the folder and no shell commands in SKILL.md (its code samples are javascript and typescript).
From the folder's file list and the shell code blocks in SKILL.md.
Links to these hosts (documentation or services it may open):
github.comFrom URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Lunora Design loads about 4.5k tokens when it runs, and up to ~9.4k if it reads all its reference files. Until then it costs about 59 tokens; SKILL.md has 2,414 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from anolilab/lunora at commit 8597a6b, republished under its MIT licence (© anolilab). 2,414 words, ~4,515 tokens.
.claude/skills/lunora-design/SKILL.md (or your agent's skills folder). This skill also uses 5 other files; get the full folder from GitHub.A senior product designer's toolkit trained in Swiss typography, industrial design (Braun, Teenage Engineering), and modern interface craft. Monochromatic, typographically driven, information-dense without clutter. Dark and light mode with equal rigor.
Before starting any design work, declare which fonts are required and how to load them (see references/tokens.md Section 1). Never assume fonts are already available.
Every screen has exactly three layers of importance. Not two, not five. Three.
| Layer | What | How |
|---|---|---|
| Primary | The ONE thing the user sees first. A number, a headline, a state. | Geist Sans at display size. --text-display. 48–96px breathing room. |
| Secondary | Supporting context. Labels, descriptions, related data. | Geist Sans at body/subheading. --text-primary. Grouped tight (8–16px) to the primary. |
| Tertiary | Metadata, navigation, system info. Visible but never competing. | Geist Mono at caption/label. --text-secondary or --text-disabled. ALL CAPS. Pushed to edges or bottom. |
The test: Squint at the screen. Can you still tell what's most important? If two things compete, one needs to shrink, fade, or move.
Common mistake: Making everything "secondary." Evenly-sized elements with even spacing = visual flatness. Be brave — make the primary absurdly large and the tertiary absurdly small. The contrast IS the hierarchy.
Per screen, use maximum:
Geist is settled, including for display. Reference sites in this space use
wider geometric grotesques, and a display face is the first thing that looks
"missing" when comparing side by side. It is not: Geist at 700 with the
negative tracking in 2.10 is the tuned setting, and --font-display /
--font-heading alias Geist deliberately rather than by omission. Do not
propose swapping the display face as a fix for a page that looks off — measure
the scale first (see 3.1), because a dropped size utility looks exactly like a
wrong typeface.
Think of it as a budget. Every additional size/weight costs visual coherence. Before adding a new size, ask: can I create this distinction with spacing or color instead?
| Decision | Size | Weight | Color |
|---|---|---|---|
| Heading vs. body | Yes | No | No |
| Label vs. value | No | No | Yes |
| Active vs. inactive nav | No | No | Yes |
| Hero number vs. unit | Yes | No | No |
| Section title vs. content | Yes | Optional | No |
Rule of thumb: If reaching for a new font-size, it's probably a spacing problem. Add distance instead.
Spacing is the primary tool for communicating relationships.
Tight (4–8px) = "These belong together" (icon + label, number + unit)
Medium (16px) = "Same group, different items" (list items, form fields)
Wide (32–48px) = "New group starts here" (section breaks)
Vast (64–96px) = "This is a new context" (hero to content, major divisions)If a divider line is needed, the spacing is probably wrong. Dividers are a symptom of insufficient spacing contrast. Use them only in data-dense lists where items are structurally identical.
Each step down adds visual weight. Use the lightest tool that works. Never box the most important element — let it float on the background.
In a monochrome system, the gray scale IS the hierarchy. Max 4 levels per screen:
--text-display (100%) → Hero numbers. One per screen.
--text-primary (90%) → Body text, primary content.
--text-secondary (60%) → Labels, captions, metadata.
--text-disabled (40%) → Disabled, timestamps, hints.The aurora ramp (cyan → violet → rose) is not part of the gray hierarchy. It's light, not paint — an event, not a default. Aurora Violet is the primary glow (the closest thing to a brand color); cyan = info/active, rose = emphasis. If >~10% of a view is aurora, pull back. Reach for the full gradient ribbon only on the focal moment (hero clause, active state, focus ring).
Emphasis is achromatic. The loudest control on a view — the primary button, the strongest interactive state — is bright, not saturated. It takes near-white on night (near-black on ivory), not the accent. Spending the accent on the primary button is the most common way a Lunora surface drifts colourful: the button is on every view, so the accent stops being an event. Keep a distinct emphasis / on-emphasis token pair for this and leave the accent for the moments below.
The accent budget, concretely. On a given view, chromatic colour is allowed in:
Everything else is grey. Anything not on that list wanting colour is asking for emphasis or ink.
Measure it, don't eyeball it. Colour creep is invisible while you add it, one component at a time. Count chromatic nodes in the rendered page and keep the ratio under ~10%:
// In the browser console on the rendered page.
const chromatic = [...document.querySelectorAll("*")].filter((el) => {
const m = getComputedStyle(el).color.match(/\d+/g);
if (!m) return false;
const [r, g, b] = m.map(Number);
return Math.max(r, g, b) - Math.min(r, g, b) > 40; // chromatic, not grey
});
console.log((chromatic.length / document.querySelectorAll("*").length) * 100);A landing page that reads "too colourful" usually measures 9-12%. The same page after moving ticks, arrows, and per-card icons to grey measures 5-6%, and the accent starts meaning something again.
Data status colors (success green, warning amber, error red) are exempt from the aurora-only rule when encoding data values. Apply color to the value itself, not labels or row backgrounds. See references/tokens.md for the full color system.
Be consistent in: Font families, label treatment (always Geist Mono ALL CAPS), spacing rhythm, color roles, component shapes, alignment.
Break the pattern in exactly ONE place per screen: An oversized number, a circular widget among rectangles, an aurora-ribbon clause among Moonlight text, a vast gap where everything else is tight.
This single break IS the design. Without it: sterile grid. With more than one: visual chaos.
Asymmetry > symmetry. Centered layouts feel generic. Favor deliberately unbalanced composition:
Balance heavy elements with more empty space, not with more heavy elements.
36GB/s in Geist Mono at 48px IS the visual. No illustrations needed.When 3+ data sections appear on one screen, vary the visual form:
| Form | Best for | Weight |
|---|---|---|
| Hero number (large Geist Sans/Geist Mono) | Single key metric | Heavy — use once |
| Segmented progress bar | Progress toward goal | Medium |
| Concentric rings / arcs | Multiple related percentages | Medium |
| Inline compact bar | Secondary metrics in rows | Light |
| Number-only with status color | Values without proportion | Lightest |
| Sparkline | Trends over time | Medium |
| Stat row (label + value) | Simple data points | Light |
Lead section → heaviest treatment. Secondary → different form. Tertiary → lightest. The FORM varies, the VOICE stays the same.
Sections 2.1–2.9 assume a product screen. Marketing surfaces carry the same voice with three additional structures. They are not decoration; each one is load-bearing, and half-applying them is what makes a page read as generic.
Page header. A full-bleed saturated colour field with a dark panel set into its lower-left, overhanging the field's bottom edge. The overhang is the composition — a panel that sits neatly inside the band reads as a banner with a box on it. Build it by pulling the panel up with a negative margin while it stays in normal flow, never by positioning it absolutely: in flow the panel's own height sets the overhang and the following content is pushed down for free, where absolute positioning needs a spacer kept in sync by hand with a height that changes per breakpoint and per content.
The field is one brand colour with depth, not three in equal measure. Violet carries it; cyan and rose are edge blooms for dimension. An even three-way split reads as a stock mesh gradient and belongs to no brand. Punch a canvas-coloured halftone dot matrix through it (≈9px grid) so it resolves from cells rather than blurring.
The navbar sits over this field. Aurora accents are mid-lightness, so dark ink on them fails contrast — keep light ink and guarantee it with a top scrim gradient rather than flipping the nav to dark.
Numbered sections. Each band opens with an index stacked over a category
label (01 / FRAMEWORK) in the left column, title and lead beside it. The
label says what kind of section it is, which the title rarely does. Index in
accent, label in faint grey.
Hairline grid. Cells separated by real 1px lines, not whitespace: the container paints the hairline colour and a 1px gap lets it through between opaque cells. Two consequences that bite:
Type scale for marketing. The 3-sizes-per-screen budget in 2.2 governs product screens. Marketing surfaces use the published scale, capped so headings stop growing on wide monitors while gutters stay fluid:
| Role | Size | Line height | Tracking |
|---|---|---|---|
| display | 3.3rem cap | 0.95 | -0.04em |
| h1 | 2.75rem cap | 1.05 | -0.04em |
| h2 | 2.5rem cap | 1.05 | -0.035em |
| h3 | 1.4375rem cap | 1.08 | -0.028em |
| body | 0.9375rem | 1.55 | -0.01em |
| blurb | 0.8125rem | 1.5 | -0.01em |
| kicker (mono, caps) | 0.6875rem | 1.2 | 0.12em |
| micro (mono, caps) | 0.625rem | 1.5 | 0.18em |
Negative tracking on display type is not optional. Large text set at default tracking is the single clearest tell of an untuned page.
Vary the layout family. A landing page that uses the hairline grid for four consecutive bands reads monotonous however good each band is. Rotate: hairline grid, rule-topped columns, a real table, a disclosure list, a full-bleed panel. No family twice in a row.
[LOADING...] text or segmented spinner.[SAVED], [ERROR: ...]rounded-none) on structural chrome — nav, console, buttons, cards, panels. Small radii (4–8px) only inside dense data (badges, table cells, chips). Never rounded glassy cards.These do not throw, do not fail a type check, and do not fail a lint. They ship a page that looks subtly wrong for reasons no one can point at.
tailwind-merge eats named type-scale utilities. When the scale is named
(text-h2, text-blurb) rather than sized (text-lg), tailwind-merge cannot
distinguish it from a colour utility. It sees cn("text-h3", "text-ink"),
assumes both are text colours, resolves the "conflict" in favour of the last
one, and drops the size. Headings then render at the browser default while
the class list still looks right in the source.
This is not hypothetical: it shipped, and every h3 on a landing page rendered
at 16px instead of 23px until it was measured.
Register the names as a font-size group, once, where cn is defined:
const twMerge = extendTailwindMerge({
extend: {
classGroups: {
"font-size": [{ text: ["display", "h1", "h2", "h3", "body", "blurb", "kicker", "micro"] }],
},
},
});Any token added to --text-* must be added here too. Verify with a direct
assertion (twMerge("text-h3", "text-ink") must keep both) rather than trusting
that it looks fine.
@theme vs @theme inline. Tokens declared in a plain @theme block are
inlined at build time, so a runtime theme swap or a [data-theme] override does
nothing. Use @theme inline when the value references another custom property,
so utilities compile to var(--…).
Tailwind namespace names are not free-form. --container-shell produces
max-w-shell, not max-w-container-shell. Check the generated utility exists
before building a layout on it; a missing utility is simply no styling, not an
error.
Verify the rendered page, not the source. Every trap above is invisible in
the JSX and obvious in getComputedStyle. Before calling a surface done, read
back the computed font sizes, line heights, and tracking of h1/h2/h3/body and
compare them to the table in 2.10.
references/tokens.md)references/tokens.md for exact valuesreferences/components.md for patternsreferences/platform-mapping.md for output conventionsFor detailed token values, component specs, and platform-specific guidance:
references/tokens.md — Fonts, type scale, color system (dark + light), spacing scale, radius/surface, motion, iconography, texture motifsreferences/components.md — Cards, buttons, inputs, lists, tables, nav, tags, segmented controls, progress bars, charts, widgets, overlays, state patternsreferences/platform-mapping.md — HTML/CSS, SwiftUI, React/Tailwind, Paper output conventionsCanonical token source. The shipped values live in marketing/design-tokens/
(tokens.css for Tailwind v4, tokens.ts for JS) with intent in
marketing/design-tokens/DESIGN.md. Those are authoritative; the references here
restate them for the methodology. When generating real code, import the tokens
rather than re-typing hex values.
Adapted under the MIT License from the "Nothing Design Skill" by Dominik Martin
(https://github.com/dominikmartn/nothing-design-skill). The craft methodology is
the original author's; the palette, typeface, corners, and accent rules were
reconciled to Lunora's established design language (marketing/design-tokens/DESIGN.md).
See LICENSE.
© anolilab, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 5 other files (references) in .agents/skills/lunora-design of anolilab/lunora.
Open the folder on GitHubat commit 8597a6b
Lunora Design next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Lunora Design this skillanolilab/lunora | 284 | — | ~4.5k | Automated safety check: Pass | MIT | |
| Impeccablebestofjs/bestofjs | 3.1k | 26 repos | ~2.6k | Automated safety check: Pass | MIT | |
| Figma Design System Builderwarpdotdev/warp | 65k | 2 repos | ~4.4k | Automated safety check: Pass | AGPL-3.0 | |
| Figma use_figma Plugin API Ruleswarpdotdev/warp | 65k | 4 repos | ~4.4k | Automated safety check: Pass | AGPL-3.0 | |
| UI StylingOhh-889/skyroc | 795 | 13 repos | ~2.5k | Automated safety check: Pass | MIT | |
| Shadcnsupabase/evals | 145 | 43 repos | ~4.5k | Automated safety check: Pass | Apache-2.0 |
bestofjs/bestofjs
A skill your agent uses when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or otherwise improve a…
warpdotdev/warp
Builds or updates a design system in Figma from a codebase in ordered phases: discovery, variables and tokens, components, theming and documentation, with checkpoints.
warpdotdev/warp
Required groundwork before any use_figma call: the rules and reference files for running JavaScript in a Figma file through the Plugin API without common failures.
Ohh-889/skyroc
Create beautiful, accessible user interfaces with shadcn/ui components (built on Radix UI + Tailwind), Tailwind CSS utility-first styling, and canvas-based visual designs.
supabase/evals
Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI.
Ohh-889/skyroc
Token architecture, component specifications, and slide generation.
anolilab/lunora
Routes general Lunora requests to the right Lunora skill and gives the shared mental model (codegen loop, generated api/internal references, review commands, add-on capabilities, the @lunora/mcp…
anolilab/lunora
Builds a reusable Lunora capability, either as a registry item that lunora add / lunora registry add copies into an app's lunora/, or as a publishable @lunora/ workspace package in the Lunora…
anolilab/lunora
Deploys a Lunora app to Cloudflare Workers + Durable Objects and gets wrangler.jsonc, remote resources and secrets to line up.
anolilab/lunora
Plans and runs Lunora schema and data migrations with the widen → migrate → narrow pattern.
anolilab/lunora
Diagnoses and fixes Lunora performance problems — full-table scans and missing indexes, expensive ctx.db.related traversals, OCC write conflicts (409 CONFLICT), live queries that re-run too often…
anolilab/lunora
Creates a new Lunora project or adds Lunora to an existing app, then gets the first schema + query/mutation round-trip running.
Categories
This skill should be used when the user explicitly says "Lunora style", "Lunora design", "/lunora-design", or directly asks to use/apply the Lunora design system. Lunora Design is an agent skill from anolilab/lunora. This skill should be used when the user explicitly says "Lunora style", "Lunora design", "/lunora-design", or directly asks to use/apply the Lunora design system.
Lunora Design fits situations like: explicitly says Lunora style; directly asks to use/apply the Lunora design system; automatically for generic UI.
Run `npx skills add anolilab/lunora --skill lunora-design -a claude-code`. Or copy the skill folder (.agents/skills/lunora-design in anolilab/lunora) into .claude/skills/lunora-design in your project. Claude Code loads it when a task matches its description.
Run `npx skills add anolilab/lunora --skill lunora-design -a codex`. Or copy the skill folder (.agents/skills/lunora-design in anolilab/lunora) into .agents/skills/lunora-design in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add anolilab/lunora --skill lunora-design -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/lunora-design, .gemini/skills/lunora-design, .github/skills/lunora-design and .opencode/skills/lunora-design in your project.
SKILL.md names no scripts, command-line tools or credentials: Lunora Design is instructions for the agent only. Its frontmatter pre-approves these tools: Read, Write, Edit, Glob, Grep.
SKILL.md names 1 domain. As links in the text: github.com. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
Lunora Design is published under the MIT licence (from the LICENSE file in the skill folder). It allows redistribution, so the full SKILL.md is shown on this page.
About 4.5k tokens (SKILL.md is roughly 18k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 4.9k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Lunora Design: Impeccable (bestofjs/bestofjs, 3.1k stars), Figma Design System Builder (warpdotdev/warp, 65k stars), Figma use_figma Plugin API Rules (warpdotdev/warp, 65k stars) and UI Styling (Ohh-889/skyroc, 795 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
anolilab (a GitHub organization) maintains it in anolilab/lunora, which has 284 GitHub stars. The repository holds 16 skills in this directory. The repository was last updated on October 11, 2026.
Source: anolilab/lunora on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.