All skills
hoodini avatar

/image-master

@83e2083

Master prompt-engineer for photoreal, artifact-free AI still images on ANY tool (Reve, Midjourney, Flux, GPT-image, Imagen, Nano Banana, Stable Diffusion). Builds prompts that hit National-Geographic-grade realism — true skin/fur texture (no plastic), correct anatomy/hands/faces, physically coherent light/shadows/reflections, clean legible text — while engineering deliberate visual IMPACT (color contrast, composition, awe/adrenaline, a striking point of view). Runs a gated process: lock the point-of-view and build the 8-block Capture Stack BEFORE generating, then run a forensic pre-submit inspection against the known artifact list. Use whenever the goal is a single still image that must look REAL and hold up under close inspection — contest entries, hero shots, product/character/wildlife/architecture/concept art, or any time AI images come out plastic, distorted, or fake. Composes with nano-banana-2 (execution) and director (motion). Hebrew triggers: תמונה, תמונות, פוטוריאליזם, ריאליזם, לייצר תמונה, פרומפט לתמונה, בלי עיוותים, עור פלסטיק, ידיים מעוותות, פרצוף מעוות, חדות, נשיונל ג'יאוגרפיק.

Use this Skill: https://skilld.dev/gh/hoodini/ai-agents-skills/image-master

This session only. Nothing lands on disk.

references13-conceptual-message-images.md

≈763 tokens on demand. Your agent reads this file only when SKILL.md points to it.

13 — Conceptual & Message-Driven Images (the idea is the hero)

The strongest contest VISION is often an IDEA, not a pretty subject. A conceptual image makes the viewer THINK and FEEL: social commentary, irony, role-reversal, juxtaposition of opposites, self-confrontation, the unexpected twist. These beat technically-pretty wildlife because most entrants do pretty — few do ideas that stop you and stay with you. This is the artist's natural voice; lean into it.

Patterns that land

  • Juxtaposition of opposites in one frame — rich/poor, free/caged, sacred/secular, human/machine, calm/danger.
  • Role reversal — animals visiting a human zoo; predator at peace with prey; the powerful made helpless.
  • Self-confrontation / doppelgänger — two versions of one person/face (poor self vs rich self).
  • Split-frame or seamless blend of two worlds/times — child/adult, prayer/battlefield, surface/depth.
  • The ironic sign or caption inside the frame — "HOMO SAPIENS"; "NO SHELTER FOR ANIMALS".
  • A punchy magazine headline that delivers the thesis — often a question: "WHO IS HAPPIER?", "PROGRESS?".

Execution rules (so the idea survives the render)

  • The emotion must be VISIBLE — the inversion read on faces and body language (the bored-rich vs the joyful-poor), not merely implied.
  • Multi-person / multi-element scenes are anatomy-risky: keep expressions candid, hands occupied or relaxed, inspect every face and hand at 100%, generate a batch and pick the clean frame.
  • Two-versions-of-one-face needs "identical facial features, the same person" stated explicitly.
  • Keep in-image SIGNS short, ALL-CAPS, English; the headline is the hero text — secondary signs are the typography risk (drop a numeric timer/long sign before the headline).
  • Validated in generation: multi-person social-commentary scenes (the human-zoo, "who is happier") render beautifully and read instantly — the voice works.

IP — DE-BRAND EVERYTHING (a contest DISQUALIFIER)

The contest requires work "without infringing on third-party intellectual property." AI defaults to rendering real logos — a KFC-style bucket, a Doritos-style bag, a Ralph Lauren polo, Apple/IBM marks, copyrighted TV-show characters. These are trademark/IP infringement and can disqualify the entry. RULE: specify "generic / unbranded / no logo" for every product, garment, device, and package; rename trademarked entities ("DEEP BLUE" → "THE MACHINE"); never depict copyrighted characters or branded show imagery. Make every reference transformative and generic. (Pop-culture "fandom battle" concepts — iPhone-vs-Android, show-vs-show — are the highest IP risk and the weakest gravitas; cut or fully abstract them.)

English text only (this artist's hard rule)

ALL in-image text is ENGLISH — Hebrew and other scripts garble in every model (proven: a shelter sign rendered "מכל מוגן"). Foreign-language or religious objects (a prayer book, a Torah) must show NO legible lettering; keep them plain/closed.

Source: SKILL.md on GitHub

No alerts3mo3 checks · Risk SAFE
  • Gen Agent Trust Hub3mo

    The skill is a comprehensive instructional framework for AI image prompt engineering. It consists entirely of Markdown documentation and does not contain any executable code, scripts, or mechanisms for data exfiltration. All external references are informative and point to legitimate AI and photography resources.

  • Socket3mo

    No alerts

  • Snyk3mo

    Risk: LOW · No issues

Signed by skilld at 83e2083. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 days ago.

Steadyupdated 4 months ago

README badge

README badge for hoodini/ai-agents-skills/image-master