Install any skill in seconds. Free to start, no credit card required.
Get Started Free →スクリーンショットに赤枠・矢印・吹き出し等の注釈を自動追加するスキル。 「スクショに注釈を付けて」「画面に矢印を追加」「マニュアル用の注釈」等のリクエストで発動。
.claude/skills/minicoohei-screenshot-annotator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 68% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 113% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -67% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -48% | 0% |
Add annotations to screenshots without modifying the original image. Annotations are overlaid on top.
bashpython scripts/annotate.py "{image_path}" "{instruction}" --style "{style}" --text "{label}" --output "{output_path}"
| Parameter | Required | Default | Description | |-----------|----------|---------|-------------| | image_path | Yes | - | Path to screenshot | | instruction | Yes | - | What to annotate (e.g., "the Login button") | | --style | No | red_box | Annotation style | | --text | No | - | Text label to add | | --output | No | auto | Output path |
| Style | Description | |-------|-------------| | red_box | Red rectangle + arrow (default) | | arrow | Red arrow pointing to element | | callout | Speech bubble with text | | highlight | Semi-transparent yellow overlay | | circle | Red circle around element | | number | Numbered marker for steps |
bash# Basic annotation python scripts/annotate.py "login.png" "the Login button" # With text label python scripts/annotate.py "settings.png" "the gear icon" --text "Click here" # Callout style python scripts/annotate.py "form.png" "email field" --style callout --text "Enter your email"
スクリーンショットに赤枠・矢印・吹き出し・ハイライト等の注釈を自動で追加するスキルです。技術ドキュメント、ユーザーマニュアル、チュートリアル作成に最適。
| 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 | | Element not found in screenshot | instruction をより具体的に記述(例: "左上の歯車アイコン") |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,464 | 5,317 | -44% | 1 | 1 | 0% | 538 | 904 | +68% | 0 | 0 | — |
case-02 | fail→pass | 8,170 | 7,748 | -5% | 1 | 1 | 0% | 401 | 855 | +113% | 0 | 0 | — |
case-03 | fail→pass | 9,414 | 9,475 | +1% | 1 | 1 | 0% | 1,845 | 1,493 | -19% | 0 | 0 | — |
case-04 | fail→pass | 14,958 | 2,772 | -81% | 1 | 1 | 0% | 2,358 | 789 | -67% | 0 | 0 | — |
case-05 | fail→pass | 8,180 | 1,855 | -77% | 1 | 1 | 0% | 1,592 | 828 | -48% | 0 | 0 | — |
case-06 | fail→pass | 10,953 | 9,880 | -10% | 1 | 1 | 0% | 1,432 | 967 | -32% | 0 | 0 | — |
case-07 | fail→fail | 5,042 | 10,294 | +104% | 1 | 1 | 0% | 174 | 1,015 | +483% | 0 | 0 | — |
case-21 | pass→pass | 11,267 | 6,127 | -46% | 1 | 1 | 0% | 1,947 | 1,522 | -22% | 0 | 0 | — |
case-08 | fail→fail | 11,044 | 5,399 | -51% | 1 | 1 | 0% | 2,092 | 842 | -60% | 0 | 0 | — |
case-09 | pass→pass | 9,960 | 1,688 | -83% | 1 | 1 | 0% | 1,759 | 821 | -53% | 0 | 0 | — |
case-10 | pass→pass | 8,875 | 3,492 | -61% | 1 | 1 | 0% | 1,645 | 1,284 | -22% | 0 | 0 | — |
case-11 | pass→pass | 8,660 | 16,076 | +86% | 1 | 1 | 0% | 1,456 | 1,687 | +16% | 0 | 0 | — |
case-12 | fail→pass | 5,628 | 1,931 | -66% | 1 | 1 | 0% | 825 | 793 | -4% | 0 | 0 | — |
case-13 | fail→fail | 4,838 | 5,424 | +12% | 1 | 1 | 0% | 259 | 861 | +232% | 0 | 0 | — |
case-14 | fail→pass | 36,369 | 4,326 | -88% | 1 | 1 | 0% | 5,882 | 803 | -86% | 0 | 0 | — |
case-15 | fail→pass | 1,923 | 1,820 | -5% | 1 | 1 | 0% | 244 | 807 | +231% | 0 | 0 | — |
case-22 | pass→pass | 9,721 | 7,999 | -18% | 1 | 1 | 0% | 1,693 | 2,019 | +19% | 0 | 0 | — |
case-16 | pass→pass | 8,119 | 1,973 | -76% | 1 | 1 | 0% | 1,144 | 890 | -22% | 0 | 0 | — |
case-17 | fail→pass | 2,994 | 1,434 | -52% | 1 | 1 | 0% | 437 | 771 | +76% | 0 | 0 | — |
case-18 | fail→pass | 20,290 | 1,317 | -94% | 1 | 1 | 0% | 3,371 | 707 | -79% | 0 | 0 | — |
case-19 | fail→fail | 5,668 | 5,423 | -4% | 1 | 1 | 0% | 257 | 851 | +231% | 0 | 0 | — |
case-20 | pass→pass | 7,514 | 5,203 | -31% | 1 | 1 | 0% | 1,518 | 1,552 | +2% | 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 16 counted toward the lift figure. The other 6 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 +50 percentage points is the difference between those two pass rates over the 16 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.