Agent skill

Beautify With Pingfusi

by alex-durango in alex-durango/pingfusi

Beautify or redesign an existing website through iterative pingfusi review rounds with a real human reviewer.

MITAuto-check passedFrontend & Design

Install Beautify With Pingfusi

skills CLI
$ npx skills add alex-durango/pingfusi --skill beautify-with-pingfusi -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install alex-durango/pingfusi beautify-with-pingfusi --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/alex-durango/pingfusi.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skill/beautify-with-pingfusi .claude/skills/beautify-with-pingfusi && rm -rf skills-src

Use ~/.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/

Facts

Skill name
beautify-with-pingfusi
GitHub stars
125
Token cost
~1.7k tokens
SKILL.md length
636 words
Files
2
Skills in repo
5
Repo updated
First seen
Licence
MIT

At a glance

Beautify or redesign an existing website through iterative pingfusi review rounds with a real human reviewer.

  • Works in 8 steps: Run pingfusi doctor. Locate the real… → Build the untouched page. Publish that… → Improve the page in its own source. Work… → …
  • Asked to beautify this website
  • SKILL.md covers Non-negotiables and Workflow
  • Calls npm

What it does

Beautify With Pingfusi is an agent skill from alex-durango/pingfusi. Beautify or redesign an existing website through iterative pingfusi review rounds with a real human reviewer. Use when asked to "beautify this website," "make this page look professional," "polish this UI/design," "improve the visual design," or finish an AI-built page when there is no reference site to match. Do not use for pixel-matching a known original; use fix-with-pingfusi for that.

Its SKILL.md is about 1.7k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `agents/openai.yaml`).

It sits in Frontend & Design, covering UI design. It works with Model Context Protocol. The repository describes itself as: MCP server + CLI that puts a real human in your coding agent's loop. It publishes work mid-task, a reviewer pins what's wrong and returns a verdict, and the agent iterates until… The licence is MIT.

When your agent uses it

  • Asked to beautify this website
  • Make this page look professional
  • Polish this UI/design
  • Improve the visual design

Example prompts

  • “beautify this website,”
  • “make this page look professional,”
  • “polish this UI/design,”
  • “/beautify-with-pingfusi”

Workflow steps

8 steps, taken from the first numbered list in SKILL.md.

  1. Run pingfusi doctor. Locate the real source, build/test commands, responsive
  2. Build the untouched page. Publish that static output through Pingfusi hosting and
  3. Improve the page in its own source. Work in this order: hierarchy and composition;
  4. Build and test at desktop and phone widths. Publish the current static output through
  5. File one custom round against the current public URL. Keep it within the service caps
  6. The filing command automatically chains client-safe wait legs until feedback. If a
  7. Fetch fresh with core.review.verify(stateFile). Read every structured comment and
  8. After approval, rerun the project's tests and capture the after screenshot at exactly

What it can do on your machine

Read from SKILL.md and the folder at commit cec753b. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • npm

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use npm, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Beautify With Pingfusi loads about 1.7k tokens when it runs. Until then it costs about 104 tokens; SKILL.md has 636 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~104
When it runs · the whole SKILL.md, loaded when a task matches
~1.7k

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.

Safety

Auto-check passed

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.

SKILL.md

The full file from alex-durango/pingfusi at commit cec753b, republished under its MIT licence (© alex-durango). 636 words, ~1,717 tokens.

Download SKILL.mdSave it as .claude/skills/beautify-with-pingfusi/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
beautify-with-pingfusi
description
Beautify or redesign an existing website through iterative pingfusi review rounds with a real human reviewer. Use when asked to "beautify this website," "make this page look professional," "polish this UI/design," "improve the visual design," or finish an AI-built page when there is no reference site to match. Do not use for pixel-matching a known original; use fix-with-pingfusi for that.

Beautify a website with pingfusi

Turn an existing page into a professionally designed version, then keep iterating on element-pinned feedback until a human reviewer approves it. Treat the page's purpose, content, brand, and behavior as constraints; there is no ground-truth design to copy.

Non-negotiables

  • Edit the project's own components and styles. Preserve product intent, working behavior, semantic structure, accessibility, and user-provided copy unless the user asked to change them.
  • Freeze the exact pre-edit build before changing source. Keep its public draft record and same-viewport screenshot as the before proof; never let a mutable dev URL stand in for the baseline.
  • Publish the current build before every round. A remote reviewer cannot inspect localhost, and an unverified/dead URL burns the round.
  • File a custom single-page round with core.review.file; omit draft_url. Its reviewer surface is the current page only, with sticky comments and drawing. Do not run pingfusi review <name> file: that command is for clone fidelity and asks whether a draft is identical to an original.
  • Never approve your own work or infer approval from prose. Finish only when a fresh core.review.verify(stateFile) returns ok === true on the declared verdict.

Workflow

  1. Run pingfusi doctor. Locate the real source, build/test commands, responsive breakpoints, and any brand constraints. If review login is missing, stop and have the user run pingfusi setup; there is no offline review substitute.

  2. Build the untouched page. Publish that static output through Pingfusi hosting and save the immutable receipt, then capture a screenshot at the viewport(s) the final proof will use:

    sh
    pingfusi publish <built-dir> --name <name>-before \
      --record .pingfusi/beautify/<name>/before.json --json
  3. Improve the page in its own source. Work in this order: hierarchy and composition; typography; spacing and alignment rhythm; color, contrast, and surfaces; responsive behavior; states and finishing details. Prefer one coherent visual idea over a pile of effects. If motion materially helps, add one restrained, purposeful beat and honor prefers-reduced-motion.

  4. Build and test at desktop and phone widths. Publish the current static output through Pingfusi hosting, using a new URL for every round:

    sh
    pingfusi publish <built-dir> --name <name>-current \
      --record .pingfusi/beautify/<name>/current.json --json

    Use a verified tunnel only if the production app genuinely requires a live server and cannot produce a self-contained build. Never tunnel merely because development happens through npm run dev.

  5. File one custom round against the current public URL. Keep it within the service caps (20 steps; 300 characters per step; 40 per option). Use one result for the normal loop; use more only when the user explicitly wants higher-confidence review.

    js
    const fs = require("node:fs");
    const path = require("node:path");
    const { execFileSync } = require("node:child_process");
    const kit = execFileSync("pingfusi", ["where"], { encoding: "utf8" }).trim();
    const core = require(path.join(kit, "packages/core"));
    const stateFile = path.resolve(".pingfusi/beautify/<name>/review.json");
    fs.mkdirSync(path.dirname(stateFile), { recursive: true });
    const currentDraft = JSON.parse(fs.readFileSync(".pingfusi/beautify/<name>/current.json", "utf8"));
    const currentPublicUrl = currentDraft.url; // the just-published, verified record
    const verdicts = ["Professionally polished", "Needs another pass"];
    const { ping_id } = await core.review.file(stateFile, {
      url: currentPublicUrl,
      title: "Is this page professionally designed?",
      instructions: "Judge the current page on its own purpose. Add sticky comments or draw on exact current-page regions that still feel generic, messy, inconsistent, or hard to use.",
      steps: [
        { text: "First impression: does this feel intentionally and professionally designed?", options: ["Clearly polished", "Almost there", "Still rough"], check: null },
        { text: "Check hierarchy, typography, spacing, alignment, color, and contrast. Pin a sticky comment or draw on every current-page region that needs a specific change.", check: null },
        { text: "Check desktop and phone layouts, interaction states, and any motion. Note clipping, awkward wrapping, weak affordances, or distracting effects.", check: null },
        { text: "FINAL REQUIRED STEP — verdict. Pick one exactly.", options: verdicts, check: null },
      ],
      verdict_options: verdicts,
      approve_verdicts: [verdicts[0]],
      n_target: 1,
      // The single-page reviewer stores sticky comments and drawings directly;
      // it does not upload a reviewer screenshot.
      require_evidence: "none",
    });
  6. The filing command automatically chains client-safe wait legs until feedback. If a raw MCP leg returns pending, immediately call pingfusi_wait again; never return pending to the user or file a duplicate. The user does not request this continuation. Passive result/verify reads do not renew the lease (a lapse only pulls the round from the feed for new claims; a reviewer mid-review can still finish).

  7. Fetch fresh with core.review.verify(stateFile). Read every structured comment and selector, fix each pin in the project's own source, rebuild, republish to a new immutable current URL, and file another round. Put a concise “changed since the last review” step near the start of each refile. Each refile owns its own wait.

  8. After approval, rerun the project's tests and capture the after screenshot at exactly the before viewport(s). Keep raw round state and comments private; publish only a sanitized before/after visual and approval receipt.

Show full SKILL.md (85 more words)Show less

Use pingfusi ask only for a consequential mid-build choice with 2–3 concrete options, and only when the choice is fully expressible in words (a headline's tone, which copy angle to lead with). The ask reviewer sees only your question text — never the page or the candidates — so a visual choice (type directions, palettes, hero compositions) is never an ask: publish the variants (pingfusi publish each, or one page showing both) and file a review round instead. It is advisory and never replaces the approving round.

© alex-durango, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in skill/beautify-with-pingfusi of alex-durango/pingfusi.

  • SKILL.md
  • agents/openai.yaml

Open the folder on GitHubat commit cec753b

Compare with similar skills

Beautify With Pingfusi 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.

Beautify With Pingfusi compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Beautify With Pingfusi this skillalex-durango/pingfusi125—~1.7kAutomated safety check: PassMIT
Figma Design to Codewarpdotdev/warp65k4 repos~2.9kAutomated safety check: PassAGPL-3.0
UI UX Pro Maxsaoudi-h/solar-icons19018 repos~11kAutomated safety check: NotesCustom licence
Website Design Taste Extractorsenlindesign/taste-skill385—~5.3kAutomated safety check: PassNone
Refero Designreferodesign/refero_skill299—~5.3kAutomated safety check: PassMIT
Build Figmacursor/plugins11k—~961Automated safety check: PassNone

Similar skills

  • Figma Design to Code

    warpdotdev/warp

    Turns a Figma frame or component into production code that matches the design, using the Figma MCP server and the project's own design system.

    65k GitHub starsUsed in 4 repos~2.9k tokens
    Frontend & DesignAuto-check passed
  • UI UX Pro Max

    saoudi-h/solar-icons

    UI/UX design intelligence for web and mobile. An agent skill from saoudi-h/solar-icons.

    190 GitHub starsUsed in 18 repos~11k tokens
    Frontend & DesignAuto-check: notes
  • Website Design Taste Extractor

    senlindesign/taste-skill

    Reverse-engineers a website's design from its URL into a Markdown and JSON pair holding concrete tokens and the trade-offs behind the look.

    385 GitHub stars~5.3k tokensUpdated 3 mo ago
    Frontend & DesignAuto-check passed
  • Refero Design

    referodesign/refero_skill

    Primary/default skill for UI design, product design, web design, landing pages, dashboards, product screens, redesigns, visual polish, frontend/CSS styling, design systems, components, responsive…

    299 GitHub stars~5.3k tokensUpdated 1 mo ago
    Frontend & DesignAuto-check passed
  • Build Figma

    cursor/plugins

    Official

    Guides an agent through turning a Figma node into production UI with the repo's own components, then checks the result visually before finishing.

    11k GitHub stars~961 tokensUpdated today
    Frontend & DesignAuto-check passed
  • Better Design

    marvkr/better-design

    Build, improve, and review production interfaces with the Better Design MCP.

    254 GitHub stars~1.2k tokensUpdated 18 days ago
    Frontend & DesignAuto-check passed

More from alex-durango/pingfusi

  • Review Video With Pingfusi

    alex-durango/pingfusi

    Have any video reviewed by a real human, through iterative pingfusi review rounds.

    125 GitHub stars~1.8k tokensUpdated 1 mo ago
    Auto-check passed
  • Fix With Pingfusi

    alex-durango/pingfusi

    Fix or polish an existing website clone/draft using pingfusi review rounds.

    125 GitHub stars~1.2k tokensUpdated 1 mo ago
    Auto-check passed
  • Pingfusi Review

    alex-durango/pingfusi

    Use Pingfusi proactively whenever a coding agent reaches a question it cannot settle with code, automated tests, documentation or search, or a local browser and needs real human judgment or…

    125 GitHub stars~1.3k tokensUpdated 1 mo ago
    Auto-check passed
  • Pixel Perfect Clone

    alex-durango/pingfusi

    Clone, copy, or replicate a website/page pixel-perfect using pingfusi.

    125 GitHub stars~2.7k tokensUpdated 1 mo ago
    Auto-check passed

Questions about Beautify With Pingfusi

What does Beautify With Pingfusi do?

Beautify or redesign an existing website through iterative pingfusi review rounds with a real human reviewer. Beautify With Pingfusi is an agent skill from alex-durango/pingfusi. Beautify or redesign an existing website through iterative pingfusi review rounds with a real human reviewer.

When should I use Beautify With Pingfusi?

Beautify With Pingfusi fits situations like: asked to beautify this website; make this page look professional; polish this UI/design; improve the visual design.

How do I install Beautify With Pingfusi in Claude Code?

Run `npx skills add alex-durango/pingfusi --skill beautify-with-pingfusi -a claude-code`. Or copy the skill folder (skill/beautify-with-pingfusi in alex-durango/pingfusi) into .claude/skills/beautify-with-pingfusi in your project. Claude Code loads it when a task matches its description.

How do I install Beautify With Pingfusi in Codex?

Run `npx skills add alex-durango/pingfusi --skill beautify-with-pingfusi -a codex`. Or copy the skill folder (skill/beautify-with-pingfusi in alex-durango/pingfusi) into .agents/skills/beautify-with-pingfusi in your project. Codex loads it when a task matches its description.

Can I use Beautify With Pingfusi in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add alex-durango/pingfusi --skill beautify-with-pingfusi -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/beautify-with-pingfusi, .gemini/skills/beautify-with-pingfusi, .github/skills/beautify-with-pingfusi and .opencode/skills/beautify-with-pingfusi in your project.

What does Beautify With Pingfusi need to run?

Going by SKILL.md and its folder, Beautify With Pingfusi needs the command-line tools its instructions call (npm).

Does Beautify With Pingfusi access the network?

SKILL.md contains no URLs. Its commands use npm, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Beautify With Pingfusi safe to install?

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.

What licence does Beautify With Pingfusi use?

Beautify With Pingfusi is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Beautify With Pingfusi use?

About 1.7k tokens (SKILL.md is roughly 6.9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Beautify With Pingfusi?

Skills that share tags, products or a category with Beautify With Pingfusi: Figma Design to Code (warpdotdev/warp, 65k stars), UI UX Pro Max (saoudi-h/solar-icons, 190 stars), Website Design Taste Extractor (senlindesign/taste-skill, 385 stars) and Refero Design (referodesign/refero_skill, 299 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Beautify With Pingfusi?

alex-durango (a GitHub user) maintains it in alex-durango/pingfusi, which has 125 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on September 9, 2026.

Source: alex-durango/pingfusi on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.