메뉴
Use when scoring generated candidates for main art DNA, style-kit fidelity, gameplay role readability, Godot readiness, VFX support, and IP safety.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use when converting a gameplay/content intention into a structured asset brief before routing style kits, compiling prompts, running gacha, or scoring candidates.
Use when writing prompt, asset, review, provenance, score, and decision metadata for generated candidates and promoted assets.
Use when any asset skill needs the project art DNA, hard/soft anchors, style kits, tension dials, negative prompts, or forbidden drift rules.
Use when generating or iterating any game art asset prompt. Provides anchor-plus-aperture prompt templates for text-to-image, image-to-image, reconstruction input images, icons, UI, VFX, materials, rooms, enemies, bosses, weapons, gadgets, and asset gacha.
Use when compiling an image-to-image iteration prompt or reconstruction-ready source-image repair prompt that preserves selected strengths while mutating only chosen axes.
Use when compiling prompts for HUD icons, ability icons, cards, UI frames, panels, and small-size readable symbols.
| name | score-style-fit |
| description | Use when scoring generated candidates for main art DNA, style-kit fidelity, gameplay role readability, Godot readiness, VFX support, and IP safety. |
| version | 1.0.0 |
| author | Project Agents |
| license | Project-internal |
| metadata | {"hermes":{"tags":["evaluation","asset-review"],"related_skills":["asset-prompt-core","asset-metadata-writer"]}} |
scoring generated candidates for main art DNA, style-kit fidelity, gameplay role readability, Godot readiness, VFX support, and IP safety.
Score each 0–5: main art DNA, asset role readability, silhouette strength, creative freshness, style kit fidelity, Godot scene readiness, VFX/readability support, IP safety.
Promotion: 0–20 reject; 21–29 iterate; 30–35 shortlist; 36–40 strong but still human taste gate.
candidate_id:
score:
subscores: {}
keep_reasons: []
reject_reasons: []
required_mutations: []
promotion_decision: reject | iterate | shortlist | human_review