Skip to content
Back to search

juliobsz

MarketplaceWorks with Claude CodeGitHub

Agent skills by juliobsz

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/juliobsz--skills/badge.svg)](/listings/juliobsz--skills)

HTML

<a href="/listings/juliobsz--skills"><img src="/api/listings/juliobsz--skills/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

npx skills add wheresryan22/anatomy -g   # optional: interactive isometric explainer figures
npx skills add juliobsz/skills -g

Open a **new** agent session, then type `/awwwards` or just ask for a website.

The same repo has [/jury](#jury), which scores any site the way an Awwwards jury would: yours, one /awwwards built, or a competitor's.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 33 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. {
  2. "$schema": "https://json.schemastore.org/claude-code-marketplace.json",
  3. "name": "juliobsz",
  4. "description": "Agent skills by juliobsz",
  5. "owner": {
  6. "name": "juliobsz",
  7. "url": "https://github.com/juliobsz"
  8. },
  9. "plugins": [
  10. {
  11. "name": "awwwards",
  12. "source": "./awwwards",
  13. "description": "Build award-caliber websites from a one-line prompt: concept, art direction, motion, WebGL and a real-browser QA loop.",
  14. "author": { "name": "juliobsz", "url": "https://github.com/juliobsz" },
  15. "homepage": "https://github.com/juliobsz/skills",
  16. "repository": "https://github.com/juliobsz/skills",
  17. "license": "MIT",
  18. "category": "design",
  19. "keywords": ["web-design", "frontend", "gsap", "webgl", "awwwards"]
  20. },
  21. {
  22. "name": "jury",
  23. "source": "./jury",
  24. "description": "Score any website the way an Awwwards jury would: real-browser evidence, a blind critic, and a ranked fix list.",
  25. "author": { "name": "juliobsz", "url": "https://github.com/juliobsz" },
  26. "homepage": "https://github.com/juliobsz/skills",
  27. "repository": "https://github.com/juliobsz/skills",
  28. "license": "MIT",
  29. "category": "design",
  30. "keywords": ["web-design", "frontend", "qa", "accessibility", "performance", "awwwards"]
  31. }
  32. ]
  33. }
README.md · README · 354 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <div align="center">
  2. # /awwwards
  3. **An agent skill that builds award-caliber websites from one line.**
  4. <p>
  5. <a href="https://code.claude.com/docs/en/skills"><img src="https://img.shields.io/badge/Claude_Code-skill-1b1b19?style=flat-square&logo=claude&logoColor=white" alt="Claude Code skill"></a>
  6. <a href="https://antigravity.google"><img src="https://img.shields.io/badge/Antigravity-CLI%20%7C%20IDE%20%7C%202.0-1b1b19?style=flat-square" alt="Antigravity supported"></a>
  7. <a href="https://github.com/openai/codex"><img src="https://img.shields.io/badge/Codex-supported-1b1b19?style=flat-square" alt="Codex supported"></a>
  8. <a href="https://agentskills.io"><img src="https://img.shields.io/badge/Agent_Skills-compatible-1b1b19?style=flat-square" alt="Agent Skills compatible"></a>
  9. <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-1b1b19?style=flat-square" alt="MIT license"></a>
  10. <a href="#requirements"><img src="https://img.shields.io/badge/node-%E2%89%A5%2022.20-1b1b19?style=flat-square&logo=nodedotjs&logoColor=white" alt="Node 22.20 or later"></a>
  11. <a href="https://github.com/juliobsz/skills/actions/workflows/ci.yml"><img src="https://github.com/juliobsz/skills/actions/workflows/ci.yml/badge.svg" alt="CI"></a>
  12. </p>
  13. <p>
  14. <a href="#install"><img src="https://img.shields.io/badge/Install-npx_skills_add-d7d7d0?style=for-the-badge&labelColor=1b1b19&logo=npm&logoColor=white" alt="Install with npx skills add"></a>
  15. </p>
  16. [Demo](#demo) · [Install](#install) · [Usage](#usage) · [How it works](#how-it-works) · [Limits](#limits) · [/jury](#jury)
  17. </div>
  18. ```
  19. > make me a portfolio, my name is Julio
  20. ```
  21. That's the whole brief. The agent works as creative director, designer and creative developer
  22. in one, then hands back a site it has already checked in a real browser.
  23. ## Quick start
  24. ```bash
  25. npx skills add wheresryan22/anatomy -g # optional: interactive isometric explainer figures
  26. npx skills add juliobsz/skills -g
  27. ```
  28. Open a **new** agent session, then type `/awwwards` or just ask for a website.
  29. The same repo has [/jury](#jury), which scores any site the way an Awwwards jury would: yours,
  30. one /awwwards built, or a competitor's.
  31. ## Demo
  32. [![Watch the demo on YouTube](https://img.youtube.com/vi/BEcQoDs16IE/maxresdefault.jpg)](https://youtu.be/BEcQoDs16IE)
  33. Click on the thumbnail to watch the demo on YouTube
  34. ## Install
  35. Install with the [skills CLI](https://github.com/vercel-labs/skills). It runs through `npx`, so
  36. there's nothing to install first, and it detects your agents (Claude Code, Codex, Cursor,
  37. Gemini CLI, OpenCode, Copilot and more) and lets you pick.
  38. ```bash
  39. npx skills add juliobsz/skills -g
  40. ```
  41. | Command | Result |
  42. |---|---|
  43. | `npx skills add juliobsz/skills -g -a claude-code` | Claude Code only (use `codex`, `cursor`… for others) |
  44. | `npx skills add juliobsz/skills` | This project only, instead of your user folder |
  45. | `npx skills add juliobsz/skills -g --skill jury` | One skill only (`awwwards` or `jury`) |
  46. | `npx skills add juliobsz/skills -g --copy` | Copy the files instead of linking them |
  47. | `npx skills update -g` | Update to the latest version |
  48. | `npx skills remove awwwards jury -g` | Uninstall |
  49. The CLI collects anonymous usage data. Set `DISABLE_TELEMETRY=1` or `DO_NOT_TRACK=1` to turn
  50. it off.
  51. <details>
  52. <summary>Claude Code plugin</summary>
  53. If you prefer Claude Code's plugin system, which updates the skill for you:
  54. ```
  55. /plugin install awwwards --marketplace juliobsz/skills
  56. /plugin install jury --marketplace juliobsz/skills
  57. ```
  58. On Claude Code before 2.1.275, run `/plugin marketplace add juliobsz/skills` and then
  59. `/plugin install awwwards@juliobsz` (or `jury@juliobsz`). The skills then run as
  60. `/awwwards:awwwards` and `/jury:jury`. Use the plugin or the skills CLI, not both, or each
  61. skill shows up twice.
  62. </details>
  63. <details>
  64. <summary>Google Antigravity (CLI, IDE & 2.0)</summary>
  65. For Google Antigravity across all surfaces (CLI `agy`, Antigravity IDE, Antigravity 2.0):
  66. - **Workspace**: If working inside this repository, `.agents/skills.json` automatically registers `awwwards` and `jury`. In your own project, copy them into your `.agents/skills/` directory:
  67. ```bash
  68. mkdir -p .agents/skills
  69. cp -r <path-to-repo>/awwwards <path-to-repo>/jury .agents/skills/
  70. ```
  71. - **Global**: Copy or link the folders to your global Antigravity skills directory:
  72. ```bash
  73. mkdir -p ~/.gemini/config/skills
  74. cp -r awwwards jury ~/.gemini/config/skills/
  75. ```
  76. Start `agy`, open Antigravity IDE, or launch Antigravity 2.0 Desktop and trigger the skills via `/awwwards` or `/jury`.
  77. </details>
  78. <details>
  79. <summary>Manual install</summary>
  80. Copy the `awwwards/` folder (and `jury/`, if you want it) into your agent's skills directory,
  81. for example `~/.claude/skills/awwwards`, `~/.gemini/config/skills/awwwards` or
  82. `.agents/skills/awwwards`. Any agent that reads `SKILL.md` skills can use them. Each folder
  83. works on its own.
  84. </details>
  85. ### Works with /anatomy
  86. [/anatomy](https://github.com/wheresryan22/anatomy) by wheresryan22 is an open-source skill
  87. for detailed, interactive isometric figures that explain an idea as a real physical object.
  88. Install it next to `/awwwards` and the site can get one "how it works" section built as a real
  89. explainer figure, styled to the site's palette and grid.
  90. ```bash
  91. npx skills add wheresryan22/anatomy -g
  92. ```
  93. It's optional. Without it, `/awwwards` builds that section another way, and tells you at the
  94. end when a figure would have helped.
  95. ### Requirements
  96. - **Node 22.20+** for the skills CLI. The helper scripts need Node 20+.
  97. - **Playwright** in the site's project for screenshots and the OG image, and **axe-core** for
  98. the accessibility pass. `/awwwards` installs both as dev dependencies when it needs them.
  99. `/jury` leaves the project it reviews alone and installs them once into `~/.cache/jury`.
  100. Both use a Chromium you already have when they can.
  101. - **A model that can read images**, so the QA loop can look at its own screenshots.
  102. - **A Vercel, Netlify or Cloudflare account** only if you want it to deploy.
  103. <details>
  104. <summary>Troubleshooting</summary>
  105. - **`/awwwards` or `/jury` isn't found.** Start a new session; agents read their skills folder
  106. at startup. Then run `npx skills list -g` to see where it was installed.
  107. - **It shows up twice.** You installed it two ways (for example the skills CLI and the plugin).
  108. Remove one with `npx skills remove awwwards -g` or `/plugin uninstall awwwards@juliobsz`.
  109. - **Installed into the wrong agent.** Name the target with `-a`, for example `-a codex`.
  110. - **Screenshots fail with "Playwright not found".** Run `npm i -D playwright` in the site's
  111. project. If it reports no Chromium, run `npx playwright install chromium` once.
  112. </details>
  113. ## Why
  114. Ask an agent for "a portfolio" and you get the same site everyone gets: purple gradient, three
  115. cards, "elevate your brand". This skill replaces that default with a process.
  116. - **A concept that belongs to you.** Three ideas from different territories (your name, your
  117. craft, your place), scored on originality and fit.
  118. - **A real design system.** Palette by role, curated free fonts, grid and motion language,
  119. written to `DESIGN.md` and tested against a "genericness check".
  120. - **Motion with intent.** GSAP, Lenis and WebGL (OGL) where they earn their place, with one
  121. signature moment and everything else quiet.
  122. - **Real copy.** No lorem ipsum, fake awards or invented metrics. Anything made up is marked
  123. as a placeholder and listed at the end.
  124. - **Verified, not assumed.** A production build, a crawling audit and real-browser screenshots
  125. on desktop, mobile and reduced motion, scored against a rubric until every category hits 8/10.
  126. Its quality bar is the work that wins [Awwwards](https://www.awwwards.com) Site of the Day. The
  127. site doesn't need to be submitted anywhere; it just has to be that good.
  128. ## Usage
  129. Ask in one line and let the agent decide. It writes its assumptions down instead of running a
  130. questionnaire.
  131. ```
  132. make me a portfolio, my name is Ana
  133. a site for my architecture studio in Lisbon, we do small houses
  134. launch page for a mechanical keyboard called Kestrel, it should feel insane
  135. redesign this landing page so it could win site of the day
  136. ```
  137. Better input gives a better site. Your real name, what you do, where you are and two or three
  138. real projects let the concept come from your story instead of a placeholder.
  139. ## How it works
  140. | Phase | What happens |
  141. |---|---|
  142. | **Brief** | Infers who it's for, the audience and the site's job. Missing facts become written assumptions |
  143. | **Concept** | Three concepts from different territories (your name, your craft's materials, your place, how you work, a formal system), scored on originality and fit. The winner becomes a one-line idea with a signature moment |
  144. | **Art direction** | Palette by role, type pairing from a curated free-font library, grid, rhythm and motion language, all written to `DESIGN.md`. Then a "genericness test" kills anything an average model would have produced |
  145. | **Build** | Semantic HTML that looks finished with motion off. Then reveals, page transitions, a loader that tracks real loading, and the signature moment |
  146. | **QA loop** | Production build, a crawling audit (SEO, metadata, accessibility, weight, placeholder copy, AI-copy tells) and real-browser screenshots on desktop, mobile, reduced motion and mid-transition. It also measures LCP and CLS on a throttled phone, runs axe and tabs through every page. The agent reads every frame, a fresh critic subagent that never saw the plan scores the site against the rubric, and the agent fixes things until every category hits 8/10 |
  147. | **Ship** | OG image, favicon, sitemap, robots, then deploy to Vercel, Netlify or Cloudflare (only after you confirm) |
  148. ### Stack
  149. - **Next.js** (App Router) by default, and for anything with app features.
  150. - **Astro** for content-only showcases where motion is the product: about 75 KB of JS against
  151. about 250 KB.
  152. - **Your stack** when you run it inside an existing project.
  153. All three use [GSAP](https://gsap.com) (every plugin is free now), [Lenis](https://lenis.dev)
  154. for smooth scroll and [OGL](https://github.com/oframe/ogl) for WebGL. The motion engine is
  155. framework-agnostic TypeScript, shipped as documented code instead of a template: it runs on
  156. one clock, reverts every effect on navigation, runs transitions in parallel with the next page
  157. loading, and keeps content visible without JavaScript or with reduced motion.
  158. ### Principles it won't break
  159. - One idea, one signature moment, everything else quiet.
  160. - Every frame works as a still poster before any motion is added.
  161. - The native cursor stays. Reduced motion gets a complete, calm site. Phones get a composed
  162. layout, not a shrunken one.
  163. - No fake awards, logos or metrics. No lorem ipsum. No em dashes or "elevate your brand" copy.
  164. - No purple gradients, glassmorphism or three-card grids unless the concept truly calls for
  165. them.
  166. ## Limits
  167. - **It takes a while.** A full run is concept, design system, build and several rounds of
  168. screenshots and fixes. Expect a long session and heavy token use, not a quick snippet.
  169. - **Results depend on the model.** It was built and tested with Claude Code, Antigravity and Codex. The
  170. concept and the visual QA are only as good as the model doing them.
  171. - **Content is yours to finish.** Without real projects, photos and copy, the site ships with
  172. clearly marked placeholders that you replace.
  173. - **8/10 is its own score.** The rubric keeps it honest, but you're the final judge.
  174. ## /jury
  175. Scores any website the way an Awwwards jury would. It doesn't have to be a site /awwwards built.
  176. ```
  177. /jury https://your-site.com
  178. /jury score this project
  179. /jury https://mine.com vs https://theirs.com
  180. ```
  181. It shoots one page of every type on desktop, on a throttled phone, with reduced motion and
  182. from the keyboard. It measures LCP, CLS, FPS, axe and hover feedback, and audits SEO and
  183. metadata. Then a fresh critic that never sees the code, the design notes or your pitch scores
  184. Design, Usability, Creativity and Content, plus the six developer areas. Every number cites a
  185. screenshot or a measurement, and the fixes come ranked by how much score they'd gain.
  186. You get `.jury/<host>/scorecard.md`, next to every screenshot and report behind it. Given a
  187. local project, it builds and scores the production output, never the dev server. Re-scoring
  188. after fixes uses a new critic, so it never anchors on the old numbers.
  189. **Checked against real juries.** Run blind on two October 2026 Sites of the Day, the critic
  190. scored 6.15 and 6.10 where the jury gave 7.28 and 7.24: stricter, by a steady 1.1. So the card
  191. shows the critic's score and a jury estimate, and the band comes from the estimate. The critic
  192. is hardest on what juries rarely check (accessibility, SEO, performance on a throttled phone),
  193. which is why its fix list catches things a jury lets slide.
  194. /awwwards runs the same rubric, scripts and blind critic in its own QA loop, so a site it built
  195. and a site you built are judged the same way.
  196. - **It's an estimate.** /jury isn't affiliated with Awwwards. The criteria and weights are
  197. theirs, and public. The jury estimate rests on two sites, both winners; how it holds for
  198. weaker sites isn't measured yet.
  199. - **It judges stills and measurements.** How motion feels, and sound, stay out of the score.
  200. - **It needs a page it can reach.** Anything behind a login can't be scored; give it a local
  201. build instead.
  202. ## Standalone scripts
  203. The helpers work on any site, so you can use them on their own. The QA scripts live in
  204. `jury/scripts/`, and /awwwards carries a copy:
  205. ```bash
  206. # Path depends on your agent (e.g. ~/.gemini/config/skills/ or ~/.claude/skills/)
  207. node <skill-path>/jury/scripts/audit.mjs https://your-site.com/
  208. node <skill-path>/jury/scripts/shoot.mjs https://your-site.com/ --crawl --nav
  209. ```
  210. | Script | In | Does |
  211. |---|---|---|
  212. | `audit.mjs` | jury, awwwards | Crawl a build: SEO, a11y, weight, placeholders, AI-copy tells |
  213. | `shoot.mjs` | jury, awwwards | Screenshots of every page type: LCP/CLS on a throttled phone, FPS, overflow, tap targets, axe, keyboard focus, rest/hover pairs per pointer target, idle motion, transitions (needs Playwright; axe-core optional) |
  214. | `serve.mjs` | jury, awwwards | Static server for build output |
  215. | `og.mjs` | awwwards | Render the OG image from the live hero (needs Playwright) |
  216. | `font.mjs` | awwwards | Self-host Fontshare / Google fonts (CSS or `next/font` output) |
  217. | `photos.mjs` | awwwards | Contact sheets of free photos for marked placeholders |
  218. <details>
  219. <summary>Repository layout</summary>
  220. ```
  221. awwwards/
  222. SKILL.md the workflow and the hard rules
  223. references/
  224. quality-bar.md scoring weights, rubric, critic brief (copied from jury/)
  225. concepts.md finding an idea that can't be moved to someone else
  226. art-direction.md patterns from recent winners, archetypes, AI defaults to avoid
  227. typography.md curated free fonts with sources, scales, setting display type
  228. layout.md grids, section families, heroes, mobile composition
  229. motion.md choreography and recipes (loader handoff, iris, shared element, index preview…)
  230. webgl.md DOM-synced image planes, fullscreen shaders, performance rules (generated)
  231. patterns.md proven structures from test builds (driver module, sticky stage…)
  232. content.md voice, honest placeholders, case-study structure
  233. quality.md accessibility, performance budgets, bugs found in real builds
  234. engine.md framework-agnostic motion engine (generated)
  235. nextjs.md Next.js adapter, layout, metadata, fonts (generated)
  236. astro.md Astro adapter and layout (generated)
  237. ship.md final checks and deploy
  238. scripts/ Node helpers; shoot, audit, browser and serve are copied from jury/
  239. jury/
  240. SKILL.md the review workflow: target, evidence, blind critic, score card
  241. references/
  242. quality-bar.md scoring weights, rubric, critic brief
  243. calibration.md the critic against real juries, for the score card (never shown to the critic)
  244. scripts/ shoot, audit, browser, serve: no dependencies except Playwright (and axe-core)
  245. tests/
  246. nexttest/ Next.js app the generated references are copied from
  247. astrotest/ Astro app, same engine
  248. _gen/ templates and build.py that write the generated references and copies
  249. smoke.mjs browser smoke test run against both fixtures in CI
  250. .claude-plugin/ Claude Code plugin marketplace
  251. .github/ CI workflow, issue and PR templates
  252. ```
  253. </details>
  254. ## Security
  255. A skill is instructions your agent follows, and these have it run builds and scripts on your
  256. machine. Read each `SKILL.md` before installing, as you would any code you run. Nothing is
  257. deployed without your confirmation.
  258. ## Contributing
  259. Issues and pull requests are welcome, especially for bugs found in real builds and new
  260. patterns in `awwwards/references/`.
  261. 1. Fork the repo and create a branch.
  262. 2. Make the change. A skill is a folder with a `SKILL.md`; the skills CLI finds any top-level
  263. folder that has one. For the Claude Code plugin, also add an entry to
  264. `.claude-plugin/marketplace.json`. Engine and stack code lives in `tests/nexttest` and
  265. `tests/astrotest`, not in the markdown: edit it there and run `python tests/_gen/build.py`
  266. (see `tests/README.md`). The QA scripts and the rubric live in `jury/`, and the same command
  267. copies them into `awwwards/`, so edit them in `jury/`.
  268. 3. Test it: `npx skills add . -g -a claude-code`, then run the skill in an agent.
  269. 4. Open a pull request. The template asks what changed and what you built to check it; CI
  270. checks script syntax, `SKILL.md` frontmatter, the plugin marketplace and that the skills
  271. CLI finds every skill, on Linux, macOS and Windows. It also builds and smoke-tests both fixture
  272. apps and fails if the generated references or copies are out of date.
  273. ## Support
  274. Found a bug or have an idea? [Open an issue](https://github.com/juliobsz/skills/issues/new/choose).
  275. ## License
  276. [MIT](LICENSE) © juliobsz

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.