playbooks

home / skills / masayan1126 / masayan-uni-code-plugins / pixta-metadata

pixta-metadata skill

/plugins/gen-ai-image-prompt-creator/skills/pixta-metadata

This skill generates PIXTA-ready titles and tags from your prompt, ensuring accurate, guideline-compliant metadata for stock illustration listings.

npx playbooks add skill masayan1126/masayan-uni-code-plugins --skill pixta-metadata

Review the files below or copy the command above to add this skill to your agents.

Files (4)
SKILL.md
1.1 KB
---
name: pixta-metadata
description: PIXTAへの出品用タイトル・タグを生成。「PIXTAタグ」「PIXTA出品」「ストックイラスト登録」と言われたら使用。
---

# PIXTAメタデータ生成

## ワークフロー

1. **プロンプト受け取り**: 画像生成に使用したプロンプトを受け取る
2. **要素抽出**: プロンプトから被写体・スタイル・ポーズ・表情・背景を特定
3. **タイトル生成**: TITLE_GUIDE.md に従い3案作成（15〜30文字）
4. **タグ生成**: TAG_CATEGORIES.md の6カテゴリに沿って単語を抽出
5. **出力提示**: コピペしやすい形式で提示

## 参照ファイル

- [TITLE_GUIDE.md](TITLE_GUIDE.md): タイトル作成ルール
- [TAG_CATEGORIES.md](TAG_CATEGORIES.md): タグカテゴリ定義
- [EXAMPLES.md](EXAMPLES.md): 入出力サンプル

## 必須ルール

- タイトルは文章形式（単語羅列NG）
- タグは単語のみ（文章NG）
- プロンプト内容に忠実に（誤情報は損害賠償リスク）
- 類義語・上位概念も含める

Overview

This skill generates PIXTA-ready metadata—short titles and tag lists—for stock illustration submissions. It takes an image-generation prompt, extracts key elements, and returns three compliant title options plus organized tag keywords for quick copy-paste into PIXTA upload fields.

How this skill works

The skill analyzes the provided prompt to identify subject, style, pose, expression, and background. It then composes three sentence-style titles that follow PIXTA length and formatting rules and produces tags grouped by defined categories, including synonyms and higher-level concepts. Output is formatted for immediate use in PIXTA submission forms.

When to use it

  • Preparing metadata for PIXTA stock-illustration uploads
  • Converting an image-generation prompt into compliant titles and tags
  • Batch-processing multiple prompts to speed up submission
  • Ensuring titles are sentence-form and tags are single words
  • Avoiding rejected uploads due to poor metadata quality

Best practices

  • Provide the original image-generation prompt verbatim for best accuracy
  • Include obvious details (subject, clothing, pose, setting, mood) in the prompt
  • Verify titles for factual accuracy to avoid copyright or misrepresentation risks
  • Prefer neutral, descriptive language in titles and diverse synonyms in tags
  • Review tag list and remove any irrelevant or sensitive keywords before submission

Example use cases

  • Turn a Stable Diffusion prompt for a smiling girl in a park into three PIXTA-compliant titles and a category-organized tag set
  • Bulk-generate metadata for a series of character illustrations to speed marketplace onboarding
  • Standardize titles and tags across a contributor team to improve discoverability
  • Convert a concept sketch prompt into searchable tags including style and mood keywords

FAQ

Are titles full sentences or word lists?

Titles are sentence-form, 15–30 characters, not word lists.

Must tags be single words?

Yes. Tags are single words only; phrases are not allowed.

How faithful is the output to the prompt?

Outputs are generated to reflect the prompt exactly; verify to avoid factual errors.