More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Dashboard Design

    1

    Build, redesign and critique clean, honest analytics dashboards, with rules from 127 real ones

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Dashboard Design

    Dashboard Design

    Example session with this skill installed

    Is this a good design? [screenshot of our revenue dashboard: four KPI tiles, a revenue chart, a sources table and an insights card]

    • Read your context and instructions
    • Compiled the dashboard design
    • Generated the UI component

    Verdict

    Clean and readable, but it gives two answers to the same question. Not good yet.

    What works

    • One clear chart as the heaviest element, with the KPI strip above it.
    • The sources table is a ranked list with share bars, not a pie chart.
    • Numbers are tabular and right-aligned in the table.

    Issues, ranked by harm

    #ElementWhat the reader seesFix
    1Revenue tile vs insight cardTile says $928, the insight says "$901 this week". They count different windows.Compute once; drop the insight or make it say something new.
    2Sales deltaRed −100% because yesterday had 1 sale.Base under ~5 → grey "was 1".
    3Today vs yesterdayAt 9 a.m. today is compared with all of yesterday, so every morning looks like a crash.Compare the same elapsed time.
    4Y-axisTicks at 0, 233, 466, 699.Round ticks: 0, 250, 500, 750.
    5ColourEach tile has its own accent; the previous period is a second bright line.One accent; previous period as a grey dashed line.
    6Wide screenContent stops at 1200 px with empty bands on a 34" monitor.Let the page grow; four breakdowns in one row.

    Next steps

    I can fix these in one pass: known-totals test first (tile, chart and table each checked against a fixed dataset), then screenshots at 3440, 1440 and 375 px in light and dark before anything is committed.

    phone-menu

    1792559003821-02t2.jpg

    TSX · React component

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Gives your coding agent the taste of a senior product designer for data screens. It builds, redesigns and critiques analytics, revenue and metrics dashboards on the web (React/Tailwind) or in SwiftUI: KPI tiles, charts, comparison periods, ranked breakdowns, date range pickers, filters, insights, empty states, wide and phone layouts.

    Distilled from 127 shipped dashboard screens (analytics tools, revenue dashboards, design systems), a full teardown of one revenue-analytics product, and a real redesign where every rule was broken once, caught on a screenshot, and fixed.

    What it changes

    • A red −100% when 1 sale became 0 → a quiet grey "was 1"
    • Axis ticks like 2,103 and 1,402 → round ticks from 0 (1k, 2k)
    • Today vs all of yesterday → the same elapsed time (today until 3 pm vs yesterday until 3 pm)
    • An insight card repeating a tile → insights only when they say something new
    • A narrow column on a wide monitor → more columns as the screen grows
    • Sideways-scrolling tabs on phones → a menu grid and tabs that shrink

    How it works

    Six steps, each with a checkable "done when": name the screen's one question, lay out the hierarchy, set colour tokens (light and dark), build each component from its pattern (code snippets included), design every state (loading, stale, error, empty, tiny data), then a QA pass: known totals first (expected numbers from a fixed dataset, checked in the tile, chart and table separately), then screenshots at 3440, 1440 and 375 px. SwiftUI gets its own reference and a simulator QA pass.

    Critique mode answers "is this a good design?" with a verdict, what works, and issues ranked by harm, honesty problems first.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Frontend & Web Apps.

    Requires screenshots for effective QA. Focused on data-heavy UIs, not decorative landing pages. Optimized for specific web and SwiftUI patterns.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 27 days ago

    Software
    PlaywrightNode.js

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Browser

    The skill itself is Markdown instructions. During the QA pass the agent runs your app, takes screenshots with Playwright (web) or xcrun simctl (SwiftUI), and edits your UI files. Network access only to install Playwright if it isn't already installed.

    Listed27 days ago

    What's inside

    Frequently Asked Questions

    This skill moves beyond simple aesthetics by applying a hierarchy-first framework that prioritizes 3–6 core KPIs, ensuring data honesty through standardized comparison periods and consistent labeling of estimates.

    Web (React and Tailwind) and native SwiftUI with Swift Charts, for iPhone, iPad and Mac. The rules themselves port to any stack.

    Yes. Paste a screenshot and ask "is this a good design?". You get a one-line verdict, what works, and issues ranked by harm: numbers that disagree first, then hierarchy, then component smells, then polish, each with the fix. It can then apply the fixes in one pass.

    SKILL.md plus 8 reference files: components, layout, tokens (light and dark), code snippets, a SwiftUI reference, the QA pass, sources and a case study.

    While it can generate layouts from requirements, the design audit feature requires you to provide visual input via screenshots or live links to assess your current implementation.

    New in Frontend & Web Apps