Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    AI Image Prompt & Defect Fix Loop

    1

    This skill runs the full loop: plan, prompt, generate, inspect, fix.

    $7.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    AI Image Prompt & Defect Fix Loop

    AI Image Prompt & Defect Fix Loop

    Example session with this skill installed

    I need a product photo of a matte blue reusable water bottle on a white marble counter for an ad. I'm using Midjourney. Keep the style clean and high-end.

    • Read your context and instructions
    • Compiled the ai image prompt

    I have created a reference sheet to ensure consistency and a set of prompts optimized for Midjourney's syntax.

    Image Plan: Matte Blue Water Bottle

    REFERENCE SHEET

    Subject: Reusable water bottle, cylindrical, matte navy blue finish (#000080), silver screw-top lid, no branding or logos on the surface.

    Style: Commercial product photography, bright high-key lighting, soft shadows, 50mm lens, white marble countertop background, minimalist aesthetic.

    • Never: Plastic textures, condensation, cluttered background, vibrant neon colors.

    PROMPTS (Midjourney)
    1.

    Main: A matte navy blue reusable water bottle on a white marble countertop, silver lid, soft studio lighting, high-end commercial photography, minimalist composition, 8k resolution --ar 16:9 --v 6.0
    2.

    Variation A (Angle): Top-down bird's eye view of a matte navy blue reusable water bottle on white marble, sharp focus, clean lines, minimalist --ar 16:9 --v 6.0
    3.

    Variation B (Lighting): Matte navy blue reusable water bottle on white marble, dramatic side lighting, deep soft shadows, luxury product aesthetic --ar 16:9 --v 6.0

    NEXT STEPS

    1. Paste the Main prompt into Midjourney.
    2. Upload the resulting image here for a defect inspection.
    3. If the blue appears as a different shade, I will provide a color-weighted fix prompt.

    Note: Midjourney does not support hex codes directly, so I have used descriptive terms to guide the model toward your brand color. Check for label distortions in the silver lid area once generated.

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

    About this skill

    Image tools reward specific, ordered prompts, and the first result is rarely the keeper. This skill runs the full loop: plan, prompt, generate, inspect, fix. It can make the images with Agensi's built-in image tools, paid per image from your Agensi balance, or write prompts for your own tool.

    What it does

    • Reference sheet: locked words for subject, style and things to avoid, reused in every prompt for consistency.
    • Prompts: a main prompt and 2 variations in the order the image model reads best.
    • Makes the images: with Agensi's image tools (FLUX, Nano Banana, GPT Image). It shows the model and cost first and waits for your yes.
    • Text plan: how to get readable words in images, or exact specs to add them afterwards.
    • Defect check: hands, faces, text, logos, perspective, extra objects, cropping, product accuracy and brand colors.
    • Smallest fix: inpaint instruction, prompt change, re-roll or finish in a design tool, with the revised prompt.

    How it works

    1. Say what the image is for and its size.
    2. Approve the cost and let it generate, or paste the prompt into your own tool.
    3. Get defect notes and a fixed prompt, and run the next round.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Video Skills.

    Image generation uses Agensi's image tools and is paid per image from your Agensi balance; nothing runs without your yes. Reviewing images needs a model that can see them. It will not write prompts that copy a living artist's name, a real person's likeness or trademarked characters.

    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 10 days ago

    Permissions

    Write Files
    Read Files
    Listed10 days ago

    What's inside

    Frequently Asked Questions

    Yes, with Agensi's built-in image tools, paid per image from your Agensi balance. It shows the cost and waits for your yes first. It can also write prompts for your own tool.

    No. Agensi's image tools run on Agensi's keys. Prompts for Midjourney, Ideogram, Stable Diffusion, Firefly and similar are included if you prefer your own tool.

    It writes a locked reference sheet and reuses the exact words, and explains your tool's reference image options.

    Many tools still misspell. It keeps words short and quoted, and tells you when to add text afterwards.

    That depends on your image tool's terms; check them and any AI labelling rules where you post.

    New in Image & Video