ai image rescue and repair

    1

    Systematically diagnose and repair failed AI image prompts while preserving successful elements.

    $5

    Secure checkout via Stripe

    30-day refund guarantee

    Converts to your local currency at checkout

    0 installsSecurity scanned

    Works with the AI tools you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIGitHub CopilotGitHub CopilotGemini CLIGemini CLI+20 more

    ai image rescue and repair

    Example session with this skill installed

    My prompt 'a cozy library with a fireplace and a cat sleeping on the rug' failed. The cat has 6 toes, the books look great, and the lighting is perfect, but the fireplace is way too small.

    • Read your context and instructions
    • Compiled the ai image rescue
    • Generated the image

    I have analyzed the failure. The primary issues are a scale conflict (fireplace) and a morphological artifact (cat's paws), while the environmental assets and lighting are successful.

    Diagnosis

    • Failure class: Multi-type (Composition + Artifact)
    • Root cause: Under-specification of fireplace dimensions and model artifacting on paws.
    • Working elements (LOCK): cozy library, warm lighting, books on shelves, cat asleep.
    • Broken elements (REPAIR): fireplace size (too small), cat paw anatomy (6 toes).
    • Uncertain fragments: None.

    Rescue Prompt

    a cozy library with a fireplace and a cat sleeping on the rug, warm lighting, tall wooden bookshelves filled with books, a large stone fireplace centered on the back wall, fireplace occupies at least one-third of the frame, logs burning with visible flames, cat with natural paw anatomy, exactly 5 toes per paw, ensure all other aspects remain unchanged.
--no six toes, extra toes, tiny fireplace, bare wall, distorted paws

    Test Result

    Waiting for generation. Please provide a description of the new image once rendered.

    Evaluation

    LOCK preserved?REPAIR fixed?New issues?
    All preservedTBDTBD

    Next Step

    1. Run the prompt above.
    2. If the fireplace is corrected but the cat still has issues, I will refine only the cat fragment.
    3. If the image is perfect, we will finalize.

    ai-image-rescue-and-repair.png

    PNG · 1536×1024

    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.

    What you get

    Fix anatomical artifacts like extra fingers or warped limbs.Correct garbled or misspelled text in AI generated signage.Rescale specific objects within a composition without changing the style.Identify and resolve conflicting constraints in complex prompts.

    About this skill

    The problem

    Rewriting prompts from scratch when an AI image fails often destroys the parts you liked. You lose the composition or lighting while trying to fix a single artifact or a misspelled word.

    What it does

    • Decomposes failed prompts into "LOCK" and "REPAIR" semantic units to protect working elements.
    • Classifies failures by type including composition, identity, product accuracy, and artifacts.
    • Generates targeted rescue prompts using precise corrective syntax and negative descriptors.
    • Executes a controlled iteration loop that applies minimal deltas to reach the goal.
    • Evaluates retest results using a structured matrix to prevent regressions.

    Why this beats prompting it yourself

    Manually tweaking prompts often leads to "prompt drift" where fixing one issue introduces three new ones. This skill uses a systematic diagnostic workflow to isolate variables, ensuring you only change the specific fragments causing the failure while maintaining visual consistency.

    Use cases

    • Fixing distorted limbs or extra fingers in character portraits.
    • Correcting misspelled text on signs or product packaging.
    • Scaling architectural elements like fireplaces or windows without altering the room layout.
    • Adjusting specific lighting directions while preserving existing art styles.

    Known limitations

    Requires the user to provide the original prompt and a description of the failure. Limited to 5 iterations per rescue session.

    How to install

    Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.

    ~30 seconds
    1. 1

      Download the ZIP

      Free skills download straight away. Paid skills unlock right after purchase.

    2. 2

      Unzip into your skills folder

      Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.

    3. 3

      Ask your agent to use it

      Restart the agent if it was already running. It picks the skill up automatically - no config needed.

    Skills folder by agent

    Click the path to copy it. Create the folder if it does not exist yet.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill passes our trust checks below.

    Security scanned

    Passed our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    30-day refund

    Not a fit? Get your money back

    Trust & safety

    Security scanned

    Verified clean 21 days ago

    • Passed all security checks, Safe to install

    Listed21 days ago

    What's inside

    Frequently Asked Questions