voyager67/Pixelcraft / skills/hero-visual-repair
hero-visual-repair
VerifiedRecognizes generic AI hero visuals and rebuilds the first screen around product meaning, composition, credible imagery, and clear text placement.
Hero Visual Repair
Workflow
- Identify what the hero must communicate in the first few seconds.
- Inspect the current image, background, text safe area, product preview, CTA, and visual focal point.
- Flag generic abstract blobs, random 3D objects, stock-photo clichés, floating UI-card collages, and unrelated imagery.
- Choose a better visual from actual product output, a relevant object/environment, a functional diagram, or a restrained typographic composition.
- Place text in calm space, maintain contrast, and avoid stacking too many competing actions.
- Check crop behavior, alt text, image focal point, and mobile stacking.
- Generate imagery only when the user has requested it or approved the specific image component; do not generate a substitute for missing product evidence.
Output
Hero purpose, visual diagnosis, composition plan, image brief if needed, and responsive checks.
Non-negotiable rules
- Judge the interface that exists, not assumptions about who or what produced it. AI authorship cannot be reliably inferred from generic visual patterns alone.
- Treat “AI slop” as a shorthand for observable quality problems: generic patterns, weak hierarchy, arbitrary decoration, vague copy, fabricated-looking content, inconsistent systems, or incomplete states.
- A pattern is a signal, not proof. Record evidence, confidence, and plausible benign explanations. Do not label a screen “AI-generated” as a factual finding.
- Preserve user goals, factual content, product behavior, existing brand assets, and working accessibility unless the evidence supports a change.
- Never invent metrics, reviews, customer logos, endorsements, product claims, features, integrations, or data to make a screen look more convincing.
- Inspect rendered output when tools allow it. If you only inspect source code or a screenshot, state that limitation.
- Make the smallest effective change first. Do not replace an existing coherent visual language just to make the result look different.
- Never expose secrets or private information found in a repository. If a secret is detected, do not repeat its value; report only the file and line, stop short of publishing it, and recommend revocation/rotation.
Completion report
State the signals found, evidence and confidence, fixes made, files changed, checks actually run, and unresolved issues. Distinguish visual review, code review, and runtime testing.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |