SkillAgentSearch skills...

hallmark

Anti-AI-slop design skill for greenfield pages, audits, redesigns, and design extraction from URLs or screenshots

Install / Use

npx skills add Nutlope/hallmark --skill hallmark

Installs into whichever agent you are using.

About this skill
📄

SKILL.md

Installable skill definition

Quality Score

87/100

Supported Platforms

Universal

Our assessment of hallmark

hallmark scores 87/100 on our quality scale, 253rd of 688 AI & Machine Learning skills we index (top 37%).

Its SKILL.md is 65 KB long, well organised into 28 sections with 6 code examples: long enough that it reads more like full documentation than a focused instruction file, which agents can find harder to follow.

With 29,112 GitHub stars, it is one of the more widely adopted skills in the catalogue.

Substance
21/30
Structure
20/20
Description
12/15
Adoption
19/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated about 2 months ago, so hallmark is actively maintained.
  • It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands.

Automated pattern scan on 2026-09-26. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

hallmark compared with similar skills

All 4 of these similar skills score higher than hallmark; compare them before choosing.

SkillScoreStarsUpdatedFormat
hallmark (this skill)by Nutlope8729.1k50d agoSKILL.md
claude-memby thedotmack10094.7ktodayCLAUDE.md
Understand-Anythingby Egonex-AI10084.2k14d agoCLAUDE.md
headroomby headroomlabs-ai10073.8ktodayCLAUDE.md
CowAgentby zhayujie10047.1ktodayCLAUDE.md

Frequently asked questions

How do I install hallmark?
Run npx skills add Nutlope/hallmark --skill hallmark. The install tabs above show the steps for each supported agent.
Which AI agents does hallmark work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is hallmark safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is hallmark still maintained?
The repository was last updated about 2 months ago, so hallmark is actively maintained.

name: hallmark description: "Anti-AI-slop design skill for greenfield pages, audits, redesigns, and design extraction from URLs or screenshots. Use when the user asks to build a new app or landing page, wants to redesign something, invokes Hallmark by name, or uses audit/redesign/study." version: 1.1.0

Hallmark

A design skill for AI coding assistants. Makes the UIs they generate look made, not generated.

Hallmark is opinionated, short, and boring on purpose. It encodes a tight set of rules — drawn from the consensus of the anti-AI-slop design field (Anthropic's frontend-design skill, the Claude cookbook on frontend aesthetics, and the 2026 "tactile rebellion" movement) — and refuses to let the model fall back to the defaults every LLM was trained on.

The differentiator: Hallmark insists on structural variety, not just visual variety. Two pages by Hallmark for two different briefs should not share the same hero → 3-feature → CTA → footer rhythm. They should feel like different sites, not different colour-swaps of the same template. See references/structure.md.

Powered by Together AI.


How to use this skill

Hallmark has one default behaviour and three explicit verbs.

| Invocation | What it does | | --- | --- | | (default) | The user asked you to design or build something new. Follow the Design flow below. | | hallmark audit <target> | Read the target, score it against the anti-pattern list, return a ranked punch list. Do not edit. | | hallmark redesign <target> [--mood <name>] | Take the target's content and intent, then redesign the visual structure inside the existing implementation boundaries unless the user explicitly confirms a full rebuild. New section rhythm, new heading placement, new component voice. Preserve existing routes, component ownership, copy intent, brand, and information architecture; replace only the visual/interaction layer needed for the requested scope. | | hallmark study <screenshot \| URL> | The user pasted or attached an image of a design they admire, or pasted a URL to a live page. Extract the DNA — macrostructure, archetypes, type-pairing, colour anchor — and produce a diagnosis report, then optionally rebuild the user's content using the extracted DNA or emit a portable design.md of the DNA. Detection is automatic: a URL (http:// / https:// prefix) routes to URL mode; anything else routes to image mode. URL mode reads the page's HTML and CSS via WebFetch — it can name exact fonts and exact colour values, but can't judge rhythm. After the diagnosis, the user has three follow-ups: build with the DNA (handoff to default), lock the DNA into a portable design.md (opt-in via "lock the DNA" / "give me a design.md"), or stop at the diagnosis. Never copies pixels. Refuses template-marketplace URLs. Tighter refusal layer for design.md emission than for the diagnosis itself — URL-mode emission requires attestation that the source is the user's own or a public reference for their own brand. Falls back to asking for a screenshot if the URL is auth-walled, a JS-only SPA shell, or otherwise un-readable. Load references/study.md before this verb runs. |

If the user types anything that does not clearly map to audit, redesign, or study, treat it as default. If the user attaches an image or pastes a URL without a verb prefix, ask: "Should I study this (extract the DNA), or should I treat it as a reference for a fresh build?"

Implementation safety rail. Hallmark is a design skill, not a license to bulldoze a codebase. In any existing project:

  • Never delete production files, route trees, component directories, or an old website unless the user explicitly asks for deletion or approves a file-level plan that lists the deletions.
  • Default to in-place edits of the named files, or additive new components/tokens that are wired through the existing route. If the redesign would require removing multiple components, stop and ask for confirmation first.
  • Treat PDFs, README files, .md briefs, docs, transcripts, and pitch decks as reference material. Do not copy them word-for-word into the page unless the user explicitly says to use that text verbatim.
  • Before editing, state the exact files you expect to modify/create/delete. Deletions require explicit confirmation.

The default Design flow always picks a theme. By default it picks one of the 21 named themes — the catalog — and rotates among them per the diversification rule. There is also a quiet custom branch that constructs a one-off OKLCH palette + free-font pairing for the brief; the custom route fires only when the brief carries a creative-intent signal (the user names a brand colour, names a multi-attribute vibe the catalog can't carry, or explicitly asks for a custom theme). For vanilla briefs, the user never sees the words "catalog" or "custom" — the catalog runs silently. See Step 1 (signal detection) and Step 2.6 (dispatch); the protocol lives in references/custom-theme.md.


Disciplines that hold across every verb

These six disciplines are not verb-specific. They apply to default Design, audit, redesign, study, and component-scope alike. They sit alongside the slop test, not inside one branch of it.

  1. Pre-emit self-critique. Before handing back any output, score it 1–5 on six axes — Philosophy, Hierarchy, Execution, Specificity, Restraint, Variety. Anything < 3 triggers a revision pass. Stamp the six scores at the top of the artifact (/* Hallmark · pre-emit critique: P5 H4 E5 S4 R5 V5 */). See references/slop-test.md § Pre-emit self-critique.

  2. Honest copy — no fabricated content. If the user did not supply a metric, do not invent one. Stat-led layouts, comparison rows, and proof bars must use real numbers, a placeholder (— plus a labelled grey block, "metric to confirm"), or a different macrostructure. "+47 % conversion", "trusted by 50,000+ teams", and "10× faster" are slop the moment they're invented. Same rule for testimonials, logos, and case-study counts. See references/anti-patterns.md § Invented metrics and slop-test gate 46.

  3. Locked tokens — no mid-render improvisation. Once a theme is selected at Step 2.6, every colour and every font-family declaration in the artifact must reference a named token (var(--color-accent), font-family: var(--font-display)). Inline OKLCH / hex / rgb() values, or a font-family: "Some Font" declaration that bypasses the token block, are not allowed. If a value is needed that doesn't exist as a token, lift it into the token block as a new named variable, then reference it. See references/anti-patterns.md § Mid-render token improvisation and slop-test gate 48.

  4. Re-drawn chrome forbidden. Hallmark must not hand-build fake browser bars (URL pill + traffic-light dots), fake phone frames, fake code-block windows (mock title bar + dots wrapping a <pre>), or fake IDE chrome — the user's environment already supplies real chrome. Use real screenshots wrapped in a <figure> (with at most a hairline border), or omit the chrome and let the content stand on its own. See references/anti-patterns.md § Re-drawn UI chrome and slop-test gate 47.

  5. Mobile responsiveness — every emit verified at 320 / 375 / 414 / 768 px. Hallmark's output must render flawlessly at all four widths. The non-negotiables: no horizontal scroll + root overflow-x: clip on both html and body, never hidden (gate 34); no two-line clickable text — buttons, primary nav links, footer links, breadcrumbs, CTAs (gate 49); image-bearing grid tracks use minmax(0, 1fr), never bare 1fr (gate 50); display headers wrap inside long words via overflow-wrap: anywhere; min-width: 0 (gate 51); section heads collapse to one column on mobile across every theme variant (gate 52); radio-tab patterns don't scroll-jump (gate 53). See references/responsive.md § Mobile — non-negotiable. This is a hard floor, not a wish list.

  6. Typography purity — no italic headers. Headings and display type are always roman (font-style: normal). An italicised emphasis word inside an otherwise-upright heading (Built to <em>think</em>) is one of the most reliable AI tells; so is an all-italic display face on headings. Carry emphasis with weight, accent colour, or a drawn underline. Italic survives only as body-copy emphasis inside running paragraphs. See references/anti-patterns.md § Italic headers and slop-test gate 38a.


When the brief is a component, not a page

Before entering the full Design flow, check scope. If any of these fire, run the Component-scope flow instead — most day-to-day dev requests are component-shaped, not page-shaped, and the page-level apparatus (macrostructure, hero enrichment, footer archetype, project memory) is wrong for them.

Component-scope signals:

  • The brief names a single UI element: a button · an input · a card · a modal · a dropdown · a tooltip · a select · a checkbox · a switch · a tab strip · a chip · a badge · a banner · a snackbar · a popover · a slider · a date picker · an avatar.
  • The brief is short (≤ 30 words) and refers to one element.
  • The target file is a single component (e.g., ./Button.tsx, ./components/Input.css, app/components/Card.vue).
  • The user explicitly says "just the X", "only the Y", "this one element", "a single ___".

If two signals fire, route component. If only the page flow fires (multi-section brief, "build me a landing page"), stay in Design flow.

What Component-scope keeps from the page flow

  • Step 0 · Pre-flight scan — same. Read existing tokens, fonts, framework, microinteraction stance. A button on a Geist-bodied Tailwind project must adopt those tokens, not invent new ones.
  • Step 1 · Genre detection — same. Editorial / modern-minimal / atmospheric / playful. The component inherits its surroundings' genre (silent default to editorial when unknown).
  • Step 2.6 · Theme route — same. If a tokens.css or design.md exists, the component uses those tokens. Otherwise it asks "is there a system to follow, or should I pick one?" — defaulting to catalog if the user is silent.
  • 2+1 font discipline — same.
  • State discipline — STRICTER. Every interactive component MUST ship code for all 8 states: default · hover · :focus-visible · :active · disabled · loading · error · success. The 8-state checklist in interaction-and-states.md is mandatory, not advisory.
  • Slop test — universal-only subset. Run the visual / microinteraction / contrast (gates 40–41) / a11y / typography gates. Skip the diversification gates (no .hallmark/log.json entry — components don't rotate) and skip the layout-safety gates that assume a full page.

What Component-scope skips

  • Step 2 · Macrostructure pick. Components don't have macrostructures. State this explicitly: "Component-scope: skipping macrostructure."
  • Nav and footer archetype picks. N1a–N13 and Ft1–Ft8 are page-scope only. A component is one element; it has no nav, no footer. Skip both.
  • Hero polish patterns (HP1–HP4). Page-scope only. A button or card has no hero.
  • Step 4 · Enrichment. No hero illustration, no demo video, no abstract background. The component IS the artifact.
  • Step 5 · Multi-section preview. Replaced by the 8-state demo wrapper (below).
  • Project-memory append. No .hallmark/log.json entry for component runs. The diversification rule doesn't apply.

What Component-scope emits

Two files, side by side:

  1. The component artifact — a single self-contained file matching the project's conv

Truncated for display — read the full file on GitHub.

Related Skills

View on GitHub
GitHub Stars29.1k
CategoryAI
Updated1mo ago
Forks1.5k

Languages

CSS

Trust signals

100/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

No cautions