Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Ufo Style

ASecurity

Load when the member asks for ufo's house colours, type, spacing, radius, or other exact visual-system facts.

60 stars
0 votes
0 copies
0 views
Added 9/28/2026
ai-agentsrustgo

Works with

cli

Security Analysis

A100/100

Scanned 9/28/2026

Install to Claude Code

$npx -y skills add ufo-ai/ufo-core --skill ufo-style --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Ufo Style?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Ufo Style
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/ufo-ai-ufo-style/badge)](https://www.skillsdirectory.com/skills/ufo-ai-ufo-style)

More formats (shields.io, HTML) on the badges page.

Files
SKILL.md
---
name: ufo-style
description: "Load when the member asks for ufo's house colours, type, spacing, radius, or other exact visual-system facts."
---
# House style

The portal's own theme is the house style. Anything built for the product or for the workspace
itself takes these tokens: an application's homepage, an internal page or dashboard, a deck or a
chart the team reads. Draw from the tokens; do not invent a second palette beside them.

## When the house style is the default, and when it is not

| The build | The style |
| --- | --- |
| An application's homepage, an internal page, a dashboard, a workspace deck or chart | the house style, always |
| The member named a palette, a font, a brand, or a reference to match | theirs, whole — never a mix of the two |
| A site with an audience and a subject of its own (a public site the member is publishing) | art direction from the subject, as `website-building` step 1 orders |

The member's own direction outranks this skill wherever they gave one, including a direction given
mid-build. Say in one line which style you took and why, so a member who wanted their own can say
so before the build is finished.

## The tokens

`read` `references/tokens.css`. It is the copy-ready implementation: the eight palette declarations,
the roles drawn from them, the type scale, the spacing ramp, the radii, the easings, and the one
shadow. Copy it and `assets/fonts/` into the project with the same relative layout, import the
stylesheet first, and name only its variables afterwards — a raw hex or a raw measurement in the
page is the drift this skill exists to stop. Never fetch a house font from the network.

Paths are relative to this skill's directory (`$UFO_HOME/skills/ufo-style/`).

## What the tokens commit you to

- **Colour.** Three background steps, two text steps, the soft mark step, two accents.
  `light-dark()` carries both schemes off one declaration, so write no dark-mode variant of your own.
  Every other colour is one step or a mix of two.
- **A word takes a text step, a mark takes the mark step.** Secondary words are set in
  `--color-ink-soft`, which is the derived AA-safe pair (`#676767` light, `#A7A9A9` dark).
  `--color-mark-soft` carries `#919090` for a hairline, a rule or a dot — it measures 3.0:1 on
  the light surface, so no word is ever set in it.
- **Type.** Inter for text, Roboto Mono for code, Georgia for display, as `tokens.css` writes them.
  The portal's own display face is licensed to the portal alone: it is not in this stack, and a build
  of ours never adds it or loads it with an `@font-face` rule.
- **Size.** The scale is fixed-px chrome type, 11-24px, for app-like screens, plus one display step
  — `--text-display` (32px) — for the screen's own title, used once per screen. Nothing else on an
  app screen goes above `--text-title`. A reading or landing page keeps the fluid scale
  `website-building` teaches and takes the palette from here.
- **Chrome.** The product supplies the page's header and everything in the top right — theme,
  settings, account. A screen of ours adds no bar that crosses the centre of the page and puts no
  control in that corner; it starts at its own title.
- **Shape and motion.** One radius (`0.25rem`) for a control and a panel, `--radius-card` for a
  card or a plot, and a pill for a row.  Motion
  is short and functional: `--ease-control` for a control, the enter/leave pair for anything that
  arrives over the page.

## The mark

`assets/ufo-logo-ratio.pdf` is the logo sheet the deploy ships: the wordmark, the mark's three dots
on their golden-ratio construction, and the spacing around both. Read it before you place the mark.
When a member asks for the logo, copy the sheet into the workspace first —
`cp "$UFO_HOME/skills/ufo-style/assets/ufo-logo-ratio.pdf" /workspace/ufo-logo-ratio.pdf` — and
`share_file` that copy: `share_file` takes a path under `/workspace` alone, and this skill's tree is
not one. The file is the answer, and a copy you draw yourself is not the mark.

## A deck, a PDF, a chart

No CSS variable resolves in those, so take the hexes out of `references/tokens.css` and write them
literally. `design-foundations` carries this same palette per medium — the type sizes a format asks
for, the chart sequence, and every pairing that may carry a word — so when that skill is in the load
read its references rather than deriving any of it again.

## Before you hand it over

- No colour that is not a token or a mix of two tokens.
- Text clears WCAG AA: 4.5:1 at body size, 3:1 at 24px or at 18.66px bold. `--color-ink` clears it on
  every background step. An accent is a fill, not a word: a link is set in `--color-link`, and a
  label on top of an accent fill in `--color-fill-ink`.
- Measure the secondary pair in the light scheme and state the ratio you measured. It is the pair
  that fails: `--color-ink-soft` on `--color-surface` reads 5.4:1 and passes, the mark step in the
  same place reads 3.0:1 and does not, so measure the rendered text rather than trusting the role
  name. The dark scheme clears AA on every step (`#A7A9A9` on `#191A1A` is 7.4:1).
- Both schemes checked. The palette answers a scheme on its own, but a screenshot proves it.
- The narrow width checked at 360px, not only the desktop width: `document.scrollWidth` equals the
  viewport and no element is clipped. This is required before handover, not a suggestion. An
  application homepage is the exception: it fills one pane at one width and is checked there.

Attribution

ufo-aiufo-ai
View sourceMore from ufo-ai →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

695601 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →