More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Anti-AI Website & App Designer

    1

    Research-first design that never looks generated. Research-driven UI designer that eliminates generic AI aesthetic tells in favor of production-ready, original code.

    $9.80

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    1 use5.0 (2 reviews)
    Anti-AI Website & App Designer

    Anti-AI Website & App Designer

    Example session with this skill installed

    I need a mobile dashboard for a premium coffee roasting app. It needs to look high-end, tactile, and avoid the usual generic blue cards. Focus on the 'current roast' status.

    • Read your context and instructions
    • Compiled the anti-ai website

    I have designed the 'Current Roast' dashboard for the coffee roasting application. The design moves away from generic UI by using a high-contrast, tactile aesthetic derived from physical roasting logs and industrial equipment.

    Premium Roasting Dashboard

    Design Rationale

    Research Winner: "The Ledger" direction. I selected this because it mirrors the technical documentation used by professional roasters, moving away from soft shadows toward stark, high-contrast borders and mono-spaced data points.

    Subtractions: I removed a redundant 'Time Remaining' circular chart, a 'Roast History' card, and the bottom navigation labels to maximize focus on the live temperature readout.

    Verification: Contrast checks passed with a 14:1 ratio for the primary readout. AI-default "Inter" font was replaced with a high-contrast serif/mono pairing.

    Next steps

    • Review the tactile interaction states for the 'Adjust Airflow' button.
    • Integrate real-time WebSocket data for the temperature readout.
    • Test the parchment background visibility under direct sunlight (roastery conditions).

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

    About this skill

    🎨 Anti-AI Designer v3

    Research-First Design That Never Looks Generated

    You can spot AI-generated UI in two seconds. The same purple gradients. The same card grids. The same uppercase labels and predictable typography. The same polished-but-forgettable layouts. Different products, identical design decisions.

    Anti-AI Designer v3 gives your coding agent a structured process to break that pattern.

    Instead of jumping straight into code and relying on familiar templates, this skill requires research before visual decisions, explores three distinct creative approaches, forces a subtraction pass, and then measures the result with real scripts — not self-review.

    The objective is not simply to produce attractive UI. It is to create interfaces that feel intentional, appropriate to the product, visually distinctive, and supported by a repeatable quality-control process.

    ✨ What Makes Anti-AI Designer Different

    🔍 1. Research Before Styling

    Most generic design workflows begin with colors, gradients, cards, and components. Anti-AI Designer starts with context.

    It establishes the product’s purpose, audience, positioning, competitors, visual references, and relevant real-world materials before making major design decisions. That evidence guides the layout, typography, color palette, imagery, and overall visual language.

    Result: Design decisions connected to the actual product instead of defaults inherited from common AI-generated interfaces.

    🎯 2. Three Directions, Not One Guess

    A single design direction can lock a project into the first acceptable idea. Anti-AI Designer requires three meaningfully different creative approaches before implementation.

    • Explore distinct visual concepts rather than superficial variations of the same template
    • Consider hierarchy, composition, typography, color, density, and brand personality
    • Explain the strengths and trade-offs of each direction
    • Justify the selected approach against the product’s goals and available references

    Result: A deliberate design decision before substantial implementation work begins.

    🧹 3. A Mandatory Subtraction Pass

    Adding more visual elements does not automatically improve an interface. Unnecessary cards, decorative gradients, redundant labels, excessive borders, and generic marketing copy can make a design feel artificial and unfocused.

    Anti-AI Designer includes a dedicated subtraction pass that challenges these choices before delivery:

    • Remove decorative elements that do not support the experience
    • Replace repetitive card grids with more appropriate layouts when the content calls for it
    • Reduce unnecessary gradients, borders, shadows, and visual effects
    • Eliminate redundant headings, generic copy, and meaningless interface elements
    • Keep only the components that improve usability, hierarchy, or product identity

    Result: A more focused interface in which every visual element has a clear reason to exist.

    🛡️ 4. Script-Based Verification

    Design quality should not depend on an AI agent declaring its own work finished. The model does not get to grade its own work.

    Anti-AI Designer includes bundled Python checks that scan the finished file for common AI-design patterns and quality issues. These checks introduce a concrete verification step into the workflow rather than treating self-review as sufficient proof.

    • Scan for common design patterns targeted by the bundled checks
    • Identify relevant quality issues that need further inspection
    • Use verification results to guide corrections before delivery
    • Recheck the implementation according to the workflow

    Important: Script results are a quality gate, not a guarantee of originality or perfect design. Automated checks can identify specific patterns, but they cannot independently judge every aspect of visual quality, usability, or rendered behavior.

    ⚡ 5. Full Mode and Lite Mode

    Not every design task needs the same level of effort.

    • Full Mode: Complete research-to-verification process for new products, major redesigns, landing pages, and substantial interface projects.
    • Lite Mode: Improve an existing page, screen, or component without repeating the entire research and direction-setting process.

    Result: A structured workflow that accommodates both ambitious builds and focused improvements.

    🖥️ What Can You Use It For

    • SaaS landing pages — product-specific layouts instead of default startup templates
    • Admin dashboards — clearer information hierarchy and more purposeful data layouts
    • Web applications — consistent visual direction across product screens and workflows
    • Marketing websites — typography, imagery, layout, and brand personality aligned with the audience
    • Existing UI redesigns — identify generic patterns, remove clutter, and refine the visual experience
    • Individual components — forms, navigation, pricing sections, feature layouts, and other interface elements

    🤖 Designed for AI-Assisted Development

    Anti-AI Designer v3 works with compatible AI coding agents, including:

    • Claude Code
    • Cursor
    • Codex
    • Other skill-based coding environments

    Compatibility and available capabilities depend on the host environment, installed dependencies, and tools accessible to the agent.

    👥 Who Is It For

    • SaaS founders and indie hackers who want product interfaces that feel more considered than a default AI-generated template
    • Frontend developers who use coding agents but want more control over design decisions
    • UI/UX and product designers who want a repeatable process for guiding AI-assisted implementation
    • Agencies and freelancers who need a structured approach to producing distinctive client work
    • AI builders who want design exploration, visual refinement, and verification to be part of the development workflow

    💎 The Design Standard

    Anti-AI Designer v3 does not promise that every generated interface will be original simply because a skill was used. Instead, it establishes a more disciplined process for making design choices, challenging predictable patterns, and checking the implementation.

    The goal isn’t to make AI-generated UI prettier.
    It’s to make the design more deliberate, distinctive, and verifiable.

    Playwright-based render checks require a supported environment. Lacks network access for real-time research in some restricted environments. Focuses on UI/UX, not brand identity or logo design.

    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

    5.0
    2 reviews
    5
    2
    4
    0
    3
    0
    2
    0
    1
    0

    This skill has been used 1 time.

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    Requires

    Python

    Permissions

    Browser
    Network Access

    Allowed Hosts

    example.com
    Listed1 month ago
    Updated8 days ago

    What's inside

    Frequently Asked Questions

    Unlike standard prompts that rely on 'statistical average' layouts, this skill uses a multi-phase pipeline that enforces research-first design and systematic subtraction to remove generic AI tropes like soft-purple gradients and overused card grids.

    You will receive the core logic for the multi-phase design pipeline, custom Python verification scripts for accessibility and contrast checks, and a library of modular design token templates.

    This skill is optimized for LLMs with coding and tool-use capabilities, and it integrates seamlessly with environments supporting Python and Playwright for automated visual verification.

    The skill outputs production-ready code in HTML/CSS, React Native, Flutter, and Swift, ensuring that the visual designs are ready for immediate implementation in web or mobile projects.

    While the skill performs best with network access for real-time market research, it features a fallback mode that utilizes high-quality internal design principles if browser access is unavailable.

    New in Frontend & Web Apps