Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →PowerPointファイル（.pptx）の構造を解析し、スライド・図形・テキスト情報を出力するスキル。 「PPTXを解析」「テンプレート構造を確認」「スライドの要素を調べて」等のリクエストで発動。
.claude/skills/minicoohei-pptx-analyzer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -66% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -48% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 158% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -23% | 0% |
PowerPointファイルの構造を解析するスキル。
bash# 基本解析（JSON + TXT出力） python scripts/analyze_pptx.py template.pptx # 画像付き解析 python scripts/analyze_pptx.py template.pptx --with-images # Gemini意味解析付き python scripts/analyze_pptx.py template.pptx --with-gemini # 出力先指定 python scripts/analyze_pptx.py template.pptx --output-dir ./output
{filename}_structure.json）json{ "source_file": "template", "total_slides": 5, "slides": [ { "slide_index": 0, "layout_name": "タイトル スライド", "shapes": [ { "shape_id": 2, "name": "Title 1", "shape_type": "Shape", "left": 838200, "top": 2130425, "width": 10515600, "height": 1325563, "text": "プレゼン資料のタイトル", "has_text_frame": true, "is_placeholder": true, "placeholder_type": "TITLE (1)" } ] } ] }
{filename}_structure.txt）=== Slide 1 (Layout: タイトル スライド) ===
[2] Title 1
Type: Shape, Pos: (0.9", 2.3"), Size: 11.5" x 1.5"
Placeholder: TITLE (1)
Text: "プレゼン資料のタイトル"python-pptx: 必須Pillow: 画像処理（--with-images使用時）pdf2image + LibreOffice: PDF経由画像変換google-generativeai: Gemini解析（--with-gemini使用時）| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-13 | fail→pass | 11,609 | 2,946 | -75% | 1 | 1 | 0% | 1,875 | 1,093 | -42% | 0 | 0 | — |
case-14 | pass→pass | 10,654 | 2,056 | -81% | 1 | 1 | 0% | 1,877 | 912 | -51% | 0 | 0 | — |
case-01 | fail→pass | 16,778 | 3,561 | -79% | 1 | 1 | 0% | 3,774 | 1,283 | -66% | 0 | 0 | — |
case-02 | fail→pass | 17,862 | 7,916 | -56% | 1 | 1 | 0% | 3,782 | 1,956 | -48% | 0 | 0 | — |
case-03 | fail→fail | 16,278 | 2,852 | -82% | 1 | 1 | 0% | 2,107 | 1,112 | -47% | 0 | 0 | — |
case-04 | pass→pass | 9,394 | 2,154 | -77% | 1 | 1 | 0% | 1,863 | 851 | -54% | 0 | 0 | — |
case-05 | fail→pass | 6,959 | 3,474 | -50% | 1 | 1 | 0% | 336 | 867 | +158% | 0 | 0 | — |
case-06 | pass→pass | 9,427 | 1,921 | -80% | 1 | 1 | 0% | 1,735 | 897 | -48% | 0 | 0 | — |
case-07 | pass→pass | 7,869 | 2,719 | -65% | 1 | 1 | 0% | 1,302 | 1,102 | -15% | 0 | 0 | — |
case-08 | pass→pass | 8,647 | 3,682 | -57% | 1 | 1 | 0% | 1,578 | 1,345 | -15% | 0 | 0 | — |
case-09 | pass→pass | 11,609 | 2,854 | -75% | 1 | 1 | 0% | 1,922 | 1,177 | -39% | 0 | 0 | — |
case-10 | pass→pass | 5,557 | 2,672 | -52% | 1 | 1 | 0% | 869 | 1,038 | +19% | 0 | 0 | — |
case-11 | fail→pass | 8,063 | 2,884 | -64% | 1 | 1 | 0% | 1,417 | 1,087 | -23% | 0 | 0 | — |
case-12 | fail→pass | 7,001 | 2,075 | -70% | 1 | 1 | 0% | 1,108 | 937 | -15% | 0 | 0 | — |
case-15 | pass→pass | 4,789 | 2,358 | -51% | 1 | 1 | 0% | 651 | 980 | +51% | 0 | 0 | — |
case-16 | fail→pass | 7,463 | 2,036 | -73% | 1 | 1 | 0% | 1,196 | 910 | -24% | 0 | 0 | — |
case-17 | fail→pass | 10,111 | 1,797 | -82% | 1 | 1 | 0% | 1,650 | 892 | -46% | 0 | 0 | — |
case-18 | pass→pass | 6,906 | 1,960 | -72% | 1 | 1 | 0% | 1,076 | 893 | -17% | 0 | 0 | — |
case-19 | pass→pass | 12,315 | 12,453 | +1% | 1 | 1 | 0% | 1,906 | 2,628 | +38% | 0 | 0 | — |
case-20 | fail→fail | 11,224 | 9,182 | -18% | 1 | 1 | 0% | 2,191 | 2,571 | +17% | 0 | 0 | — |
case-21 | fail→fail | 26,801 | 38,140 | +42% | 1 | 1 | 0% | 6,172 | 6,805 | +10% | 0 | 0 | — |
case-22 | pass→pass | 6,622 | 3,798 | -43% | 1 | 1 | 0% | 1,119 | 1,255 | +12% | 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, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +36 percentage points is the difference between those two pass rates over the 21 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.