Design & Motion · v92388375-gif/pixel-style-poster-skill

pixel-style-poster

Name: pixel-style-poster
Author: v92388375-gif/pixel-style-poster-skill

Generate Pixel Style Poster prompts and matching raster images. Use when the user gives a theme, phrase, object, mood, photo, or content idea and wants a 3:4 editorial bitmap poster with a large fine dot-matrix subject, optional title-led or quiet microtype typography, close subject-text composition, optional colored-wash reverse halftone backgrounds, restrained color transitions, and low-resolution print texture. This is not retro game pixel art.

L1 Basic · Source registration onlyV1 Commit-PinnedDual-pass Audit

Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

Two-stage pure text compiler: It first compiles the user's single-line input into a 'compact, imageable, high-fidelity' final prompt, which is then rendered into a raster image using the host's built-in image generation capability. SKILL.md contains no executable code; the entire implementation relies on prompt rule orchestration.

SKILL.md

Use **Standard Mode** for all generation. Use the Fine Bitmap Prompt Compiler in this `SKILL.md` to convert the user's content into a compact, imageable, high-fidelity prompt.

Corresponds directly to the dual-output commitment stated at the beginning ('a final image-generation prompt' + 'a generated raster image made from that prompt'). The 'matching raster images' claimed in the README is actually delivered by the Codex host's built-in image generation, while the skill is solely responsible for compiling the requirements into a high-fidelity prompt.

Disciplined compilation: By default, it compiles only the parts that will become visible pixels and strictly answers rendering questions sequentially across 9 First-Principles fields (Canvas / Attention geometry / Imageable subject / Bitmap construction / Typography system / Color logic / Print texture / Emotional temperature / Hard avoids), which are then compressed into a four-paragraph Standard Prompt Shape output.

SKILL.md

Default generation should compile only the parts that become visible pixels in the final image prompt.

The catalog of 9 fields is located in SKILL.md under the 'First-Principles Prompt Fields' section ('Every Standard Mode prompt must answer these rendering questions in this order:') and converges into four sections (canvas / bitmap construction / typography+ink / scan mood+hard avoids). The documentation explicitly prevents drifting into style prose: 'Prefer a concrete, imageable prompt over a long style essay.'

Style purity maintained via two-way lists: Prefer/Avoid in Style Calibration defines an 'editorial fine bitmap print poster', while Negative Constraints and repeated anti-game pixel phrasing across engines fence in the model, pulling it back from drifting into 'retro pixel game art'.

SKILL.md

The style is editorial and print-inspired. Do not prompt it as a game asset, sprite sheet, retro arcade, Minecraft, voxel, chiptune, or 8-bit videogame screen.

Both frontmatter description and the README conclude with 'This is not retro game pixel art.' The Negative Constraints section lists 24 hard 'Always avoid' constraints (sprite, Minecraft, chibi, neon, 3D, CTA, etc.) as negative prompt material. The essence: reducing subjective 'style judgment' into an enumerable checklist of positives and negatives for the model to execute.

Diversity and anti-templating achieved through probabilistic distribution: Distribution Defaults provides 5 lanes (40/20/20/10/10), the Variation Engine supplies a six-axis variation matrix (layout / subject treatment / typography / color / texture / mood), and demands that randomness alters 'visual grammar' rather than just swapping subjects or colors.

SKILL.md

Randomness must change the visual grammar, not only the subject or color. If recent outputs used the same layout or bitmap method, choose a different one.

Each lane in Distribution Defaults has an explicit percentage and description (e.g., '40% quiet subject-led dot-matrix posters'). Single-image outputs fall back to 'If unclear, default to subject-led dot-matrix'—probabilities apply to multiple or repeated generations, while single runs prioritize contextual fit.

Closed-loop image generation and quality inspection: Workflow step 5 defaults to using the host's built-in image generation (prohibiting stopping at prompt-only) and embeds 11 error-correction branches of 'failure → regenerate once with stronger wording' (game pixel feel / pixels too coarse / over-smoothed / overly flat / sterile white background / missing title / forced large title / color wash turned blank white / subject too small / juvenile diary portrait / overly photorealistic portrait).

SKILL.md

Do not stop after prompt-only unless the user explicitly asks for prompt-only.

Example from the same section: 'If the result looks like video-game pixel art, regenerate once with stronger wording: `editorial bitmap print poster, not game pixel art, not sprite, not 8-bit game scene`.' Each rule maps an observed failure to directly appendable reinforcing phrasing, effectively transforming 'gacha-style generation' into a process governed by a finite retry strategy.

Strict output contract: Regardless of generation outcome, results are delivered in a three-section Markdown format (Generated Image + Final Prompt + Notes including Mode/Recipe/one-line interpretation), validated before delivery against a 22-item Quality Gate checklist.

SKILL.md

If generated images render directly without a file path, show the image normally and still include the final prompt.

The Output Format section defines a complete template (`**生成图** / ![Pixel Style Poster](absolute-image-path-or-rendered-image) / **最终 Prompt** / **说明** - Mode: Standard - Recipe: [layout / subject treatment / typography / color / texture / mood]`); the final item in the Quality Gate ('Did you actually generate the image unless prompt-only was requested?') reinforces the default contract of mandatory image generation.

2Core Capabilities

01Theme / mood / photo reference → dual deliverables: final image generation prompt + rendered raster image generated from that prompt

02Disciplined fixed 3:4 portrait canvas (unless the user explicitly requests an alternative aspect ratio)

039 First-Principles rendering fields answered sequentially, compiled into a four-paragraph Standard Prompt Shape

04Fine-grained negative constraint system (24 Always avoid rules + Hard Avoids fields + Style Calibration Avoids) preventing drift into game pixels, cartoons, or generic poster templates

05Probabilistic layout distribution with single-output fallback: 5 lanes (40/20/20/10/10) + six-axis variation engine

06Typography and multi-subject hierarchy discipline: microtype vs. title-led decided by input; mandatory primary/secondary visual weighting for dual/triple subjects

0711 failure-to-reinforcing-phrasing single-retry error correction branches (self-checking and self-healing loop)

08Fixed three-section Markdown output format + 22-item Quality Gate pre-delivery checklist

4Recommended Governance & Evidence

Suggested Controls for Safe Operation

For Agent Runtime

Can be auto-loaded

For Administrators

Evidence & Triggers (File Paths & Excerpts)

SKILL.md: Pure prompt directives and rule text; no script execution or external impact.

Audit Notes: Pure directive-based prompt orchestration skill: the repository contains zero executable code, executes zero network requests at runtime, accesses zero credentials, and includes zero local file read/write instructions. The only network URL is the git clone address in the README installation instructions (intended for manual developer execution, not a runtime skill action). All generation actions are delegated to the host's built-in image generation capability ('Use the built-in image generation capability by default.'), requiring no external APIs or keys. All other observations pertain purely to content-level stylistic preferences and host capability dependencies, presenting no code-level risks.

5Second-Pass Verification

[ok] frontmatter 直采：name=pixel-style-poster-skill、description 原文（含收尾 'This is not retro game pixel art.'） — raw 重读文件头 5 行逐字核对无出入。注意：frontmatter name 与任务入库名 pixel-style-poster 不一致（任务短名），README 用法示例与调用符为 $pixel-style-poster-skill；按约定 Analysis 文件以任务名入库，skill.name 保留真实 frontmatter 名。
[ok] 『生成匹配的 raster 图像』承诺是否夸大：README Output 列 1. the generated poster image — SKILL.md 明确 'Use the built-in image generation capability by default.'，图由宿主内置能力生成而非仓库代码渲染——依赖 Codex 具备图像生成，属前提明确的能力委托而非夸大；宿主缺该能力时按 'prompt-only' 例外降级（同样有明示）。
[ok] 概率车道与再生成纠错分支的数量/内容 — raw 重读 Distribution Defaults：5 条车道 40/20/20/10/10 与回退句逐字相符；Workflow step 5 逐条点数恰 11 条 'regenerate once' 分支（游戏像素/像素过粗/过平滑/过平板/死白底/标题缺失/被迫大标题/水洗变白纸/主体过小/日记人像幼稚/人像过写实），无夸大。
[ok] 安全面反查：是否存在漏掉的网络调用/隐藏脚本/混淆内容 — git ls-tree 全量 + 隐藏文件检查无脚本类文件；token 扫描（http/curl/bash/python/api[_-]key/token/secret/.env 等）唯一命中为 README 安装段 https://github.com clone URL；SKILL.md 纯 markdown 无代码块以外的可执行面。结论 green 成立。
[ok] 示例图与 3:4 画布宣称的一致性 — 3 张示例 PNG（README examples 表格 alt 名 Mint Night Lily / Rococo Velvet Grapes / Deep Teal Butterfly）元数据均 1086×1448=3:4，与 SKILL.md 'vertical 3:4' 硬要求相符。
[ok] 外部依赖清单为空是否属实（该技能靠什么成像） — 仓库无 CLI/API/网络端点/第三方包引用；成像唯一依赖是宿主内置图像生成（'built-in image generation capability'），属宿主能力非仓库声明的外部依赖，故 external_deps=[] 如实，机制已在 how_it_works/security 注明。

6Verdict & Scenarios

Engineers 'style' into executable rules: 9-field compiler + four-paragraph Prompt Shape + 24 negative constraints yield stable, imageable, high-fidelity prompts instead of vague style essays.

Self-correcting closed loop: 11 'failure observation → regenerate once with fixed reinforcing phrasing' branches converge probabilistic gacha image generation into a finite retry workflow, supplemented by a 22-item Quality Gate self-audit.

Zero dependencies and zero side effects: no scripts, no network requests, no credentials, and no file operations; the entire repository acts as the skill directory, ready upon clone, and portable across sessions.

Explicit anti-templating architecture: percentage-based distribution lanes (40/20/20/10/10) + six-axis variation engine ensure randomness shifts visual grammar rather than merely swapping colors.

Authentic proof via sample deliverables: three 1086×1448 sample images feature exact 3:4 aspect ratios, with style characteristics (dot-matrix subjects, microtype typography, scanned paper textures) fully verifiable against claims rather than empty promises.

Best for: Ideal for Codex environments with built-in image generation capabilities to rapidly transform a single theme, mood, or phrase (including Chinese requests and mixed Chinese-English copy) into stylistically consistent 3:4 fine bitmap poster prompts and final rendered images; well-suited for poster/print-style design practice, botanical/dot-matrix visual experiments, and batch image generation for layout screening; also serves as a strong reference for skill authors looking to engineer style rules via positive/negative checklists and error-correction branches.

Not for: Not suitable for host environments lacking built-in image generation (which will only output prompts); not suitable for tasks requiring strict photorealism or fidelity (the entire pipeline is explicitly anti-photorealistic, enforcing bitmap stylization and aesthetic abstraction); not suitable for producing vector artwork, high-resolution large-format print files, or editable source assets (PSD/SVG/AI); not suitable for strict non-poster canvas aspect ratios (strictly defaults to 3:4); not suitable for batch pipelines requiring strict determinism and reproducibility (due to probabilistic lane distribution).

Install Agent-Ready

① Verified Mirror 2026-09-09

Option A · Direct DownloadDownload pixel-style-poster.tar.gz

sha256: 55f77756bc3db579…

Option B · Copy Guide for Agent

install.json

Agents automatically download and verify mirrors from this JSON guide.

② Upstream GitHub

GitHub Repo ↗

Pinned to commit b93066b52f.