Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Annotated Wireframe

    2

    Draw a low-fidelity screen sketch with orange callouts, export it to PNG, and write the build prompt that goes with it.

    Free

    1 useSecurity scanned
    Annotated Wireframe

    Annotated Wireframe

    Example session with this skill installed

    Sketch the home screen of my expenses app before we build it. Show this month's total, filters for all and unpaid, a list of expenses with a mark as paid button, and an add button.

    • Read your context and instructions
    • Compiled the annotated wireframe
    • Generated the UI component

    I drew the wireframe, exported it, checked the image, and wrote the build prompt. The screenshot on this listing is this PNG, from a real run of the export script.

    Image: mockups/expenses-home/01-home.png

    Build prompt: mockups/expenses-home/prompt.md

    Task: build the screen Expenses home in Expenses web (React + Vite), at /expenses. The layout is in 01-home.png.
    
    Parts. Each orange label in the image is one line here:
    Open account settings, where do I change my account?, opens the account page
    This month's total, how much did I spend?, no click
    Filter: all, what did I spend on?, shows every row
    Filter: unpaid, what is still owed?, shows unpaid rows
    Expense name, what is this row?, opens the expense
    Date and category, when and what kind?, no click
    Mark as paid, is this settled?, marks the row as paid
    Add an expense, how do I record a new one?, opens the new expense form
    
    Data. Use these fields only:
    GET /api/expenses → name, date, category, amount, paid
    
    Use what exists: Card, Pill, Button, useExpenses
    Handle these states: Empty: "No expenses yet"
    Do not build: sign-in, editing
    
    First list the files you will change, and any part of the image that is unclear to you. Then build.
    At the end, tell me how to check the result: open /expenses at 500 px and 1280 px
    

    Attach the PNG and the prompt to your coding agent to build the screen.

    01-home.png

    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

    Give your coding agent a picture, not a paragraph. This skill draws a low-fidelity screen sketch with orange callouts, exports it to PNG, and writes the build prompt that goes with it.

    What you get

    • An HTML template with phone and browser frames, and grey building blocks
    • A script that exports the wireframe to PNG with headless Chrome
    • A build prompt with fields for data, reuse, states and scope

    How it works

    1. The agent lists the parts of the screen. Each label says what the part does for the user.
    2. The agent builds the wireframe from the template blocks.
    3. The script exports the PNG. The agent opens it and checks it against a list: each arrow ends on its part, no labels overlap, nothing is cut off.
    4. The agent fills the build prompt and saves it next to the image.

    Limits

    • It draws layouts, not finished designs.
    • Keep to about 10 callouts on one screen. Split a larger screen into two wireframes.
    • The agent does the visual check. The script only exports the image.

    Needs Python 3 and Chrome or Chromium for the PNG export. Set CHROME if the browser is not found. The export script runs on macOS and Linux, and we tested it on macOS only. It draws layouts, not finished designs. Keep to about 10 callouts on one screen.

    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 any payment.

    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

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 11 days ago

    • Free with an Agensi account

    Requires

    Python 3
    Google Chrome or Chromium

    Permissions

    Terminal / Shell
    Write Files
    Environment Variables
    Read Files
    Browser

    File Scopes

    mockups/**
    Listed11 days ago

    What's inside

    Frequently Asked Questions

    An image shows your agent where each part goes. The orange labels say what each part does. The build prompt adds what an image cannot show: data, reuse, states and scope.

    SKILL.md with the steps, an HTML template, a Python script that exports PNG files, and a build prompt template.

    Python 3, and Chrome or Chromium on macOS or Linux. Set CHROME to the browser's path if the script does not find it. The skill makes no network call.

    Keep to about 10 callouts. More than that gets hard to read, so split the screen into two wireframes and two prompts.

    Yes. The prompt is plain text and the wireframe is a PNG. The coding agent must be able to read an attached image.

    New in Frontend & Web Apps