Install any skill in seconds. Free to start, no credit card required.
Get Started Free →各種SNS・広告プラットフォーム向けのバナー/クリエイティブを生成するスキル。 X, Facebook, Instagram, PRTimes, YouTube, LINE, Web広告に対応。 「バナーを作って」「広告画像を生成」「SNS用の画像」「クリエイティブ制作」等のリクエストで発動。
.claude/skills/minicoohei-banner-creator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -14% | 0% |
各種SNS・広告プラットフォーム向けのバナー/クリエイティブを生成します。
| Platform | Size | Aspect Ratio | |----------|------|--------------| | x_post | 1200x675 | 16:9 | | x_card | 800x418 | 1.91:1 | | facebook | 1200x630 | 1.91:1 | | facebook_story | 1080x1920 | 9:16 | | instagram_feed | 1080x1080 | 1:1 | | instagram_story | 1080x1920 | 9:16 | | prtimes | 1200x630 | 1.91:1 | | youtube | 1280x720 | 16:9 | | line | 1040x1040 | 1:1 | | web_horizontal | 1200x628 | 1.91:1 | | web_vertical | 300x600 | 1:2 |
bash# Basic usage python scripts/banner_creator.py --platform x_post --message "キャッチコピー" # With copy text generation python scripts/banner_creator.py --platform instagram_feed --message "新商品発売" --with-copy # With reference image search python scripts/banner_creator.py --platform facebook --message "セール" --search-ref "EC セール バナー" # Full options python scripts/banner_creator.py \ --platform x_post \ --message "メインメッセージ" \ --sub-copy "サブコピー" \ --cta "今すぐ登録" \ --tone professional \ --color-scheme cool \ --font-style bold \ --priority ctr \ --brand-name "会社名" \ --session "campaign_name" \ --with-copy \ --variants 3
| Parameter | Required | Default | Description | |-----------|----------|---------|-------------| | --platform | Yes | - | Target platform (see table above) | | --message | Yes | - | Main headline/catchphrase | | --sub-copy | No | - | Sub-headline or details | | --cta | No | - | Call-to-action text | | --tone | No | professional | Tone: professional, casual, pop, elegant, urgent, minimal, tech, natural | | --color-scheme | No | auto | Color: warm, cool, mono, pastel, vivid, dark, or HEX code | | --font-style | No | auto | Font: gothic, mincho, handwritten, bold, script, geometric | | --priority | No | ctr | Focus: ctr, brand, info, emotion, product, event | | --brand-name | No | - | Brand/company name to display | | --reference | No | - | Local path or URL to reference image | | --search-ref | No | - | Keywords to search for reference images | | --session | No | - | Session name for organizing output | | --with-copy | No | false | Generate copy text along with image | | --variants | No | 1 | Number of variations to generate | | --output | No | auto | Output file path |
docs/generated/banners/{date}_{session}/{filename}.png{filename}_copy.mdbashpython scripts/banner_creator.py \ --platform x_post \ --message "AI時代の働き方改革" \ --sub-copy "無料ウェビナー開催" \ --cta "今すぐ登録" \ --tone professional \ --with-copy
bashpython scripts/banner_creator.py \ --platform instagram_feed \ --message "Summer Collection" \ --tone pop \ --color-scheme vivid \ --search-ref "fashion summer sale instagram"
bashpython scripts/banner_creator.py \ --platform prtimes \ --message "新サービスリリースのお知らせ" \ --brand-name "株式会社〇〇" \ --tone professional \ --priority info
各種SNS・広告プラットフォーム(X, Facebook, Instagram, PRTimes, YouTube, LINE, Web広告)向けのバナー/クリエイティブを、Gemini Image Generation APIで自動生成するスキルです。プラットフォーム別プリセット、参考画像検索、コピーテキスト同時生成に対応。
| エラー | 解決方法 | |--------|---------| | API key not found | GEMINI_API_KEY または GOOGLE_API_KEY を環境変数に設定 | | Image generation failed | Gemini API のレート制限の可能性。数秒待ってリトライ | | Unknown platform | --platform に正しいプリセット名を指定(x_post, instagram_feed 等) |
docs/generated/banners/ 以下に保存されている--with-copy 指定時にコピーテキスト .md ファイルも生成されている| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 12,035 | 2,690 | -78% | 1 | 1 | 0% | 1,747 | 1,782 | +2% | 0 | 0 | — |
case-02 | fail→pass | 18,021 | 2,423 | -87% | 1 | 1 | 0% | 2,799 | 1,867 | -33% | 0 | 0 | — |
case-03 | fail→pass | 23,664 | 3,038 | -87% | 1 | 1 | 0% | 4,115 | 1,924 | -53% | 0 | 0 | — |
case-04 | fail→pass | 13,683 | 3,150 | -77% | 1 | 1 | 0% | 2,035 | 1,774 | -13% | 0 | 0 | — |
case-10 | fail→pass | 14,996 | 5,975 | -60% | 1 | 1 | 0% | 2,223 | 1,903 | -14% | 0 | 0 | — |
case-05 | fail→pass | 12,282 | 2,762 | -78% | 1 | 1 | 0% | 1,970 | 1,837 | -7% | 0 | 0 | — |
case-06 | fail→pass | 13,428 | 2,845 | -79% | 1 | 1 | 0% | 2,069 | 1,798 | -13% | 0 | 0 | — |
case-07 | fail→pass | 17,841 | 3,563 | -80% | 1 | 1 | 0% | 2,684 | 1,871 | -30% | 0 | 0 | — |
case-08 | fail→pass | 25,041 | 2,479 | -90% | 1 | 1 | 0% | 2,467 | 1,761 | -29% | 0 | 0 | — |
case-09 | fail→pass | 29,292 | 2,422 | -92% | 1 | 1 | 0% | 3,041 | 1,713 | -44% | 0 | 0 | — |
case-11 | fail→pass | 15,187 | 3,038 | -80% | 1 | 1 | 0% | 2,236 | 1,857 | -17% | 0 | 0 | — |
case-12 | fail→fail | 19,436 | 3,200 | -84% | 1 | 1 | 0% | 2,274 | 1,922 | -15% | 0 | 0 | — |
case-13 | fail→pass | 14,746 | 4,412 | -70% | 1 | 1 | 0% | 2,197 | 2,108 | -4% | 0 | 0 | — |
case-14 | fail→pass | 12,023 | 1,955 | -84% | 1 | 1 | 0% | 1,946 | 1,748 | -10% | 0 | 0 | — |
case-15 | fail→pass | 14,934 | 2,126 | -86% | 1 | 1 | 0% | 2,211 | 1,744 | -21% | 0 | 0 | — |
case-16 | fail→pass | 14,776 | 2,698 | -82% | 1 | 1 | 0% | 2,203 | 1,833 | -17% | 0 | 0 | — |
case-17 | fail→pass | 16,565 | 2,252 | -86% | 1 | 1 | 0% | 3,016 | 1,750 | -42% | 0 | 0 | — |
case-18 | fail→pass | 14,221 | 2,968 | -79% | 1 | 1 | 0% | 2,006 | 1,748 | -13% | 0 | 0 | — |
case-19 | fail→fail | 15,086 | 5,390 | -64% | 1 | 1 | 0% | 2,317 | 1,939 | -16% | 0 | 0 | — |
case-20 | pass→pass | 17,071 | 11,918 | -30% | 1 | 1 | 0% | 3,567 | 3,887 | +9% | 0 | 0 | — |
case-21 | pass→pass | 28,528 | 8,434 | -70% | 1 | 1 | 0% | 2,923 | 2,836 | -3% | 0 | 0 | — |
case-22 | pass→pass | 7,633 | 11,857 | +55% | 1 | 1 | 0% | 1,156 | 3,166 | +174% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +77 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.