Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →CSVファイルの行数・列数の取得、データ型推定、欠損値検出、数値列の統計情報を出力するスキル。 「CSVを分析して」「CSVの中身を確認」「データの概要を見せて」等のリクエストで発動。
.claude/skills/minicoohei-csv-analyzer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -61% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -63% | 0% |
「CSV分析」「CSVファイル解析」「データ概要」「欠損値チェック」「CSV統計」
CSVファイルを分析し、統計情報とデータ型推定を行うSkillです。
bashpython skills/csv-analyzer/scripts/analyzer.py --input data.csv
pythonfrom scripts.analyzer import CSVAnalyzer analyzer = CSVAnalyzer("data.csv") result = analyzer.analyze() print(result)
json{ "filename": "data.csv", "rows": 1000, "columns": 5, "encoding": "utf-8", "file_size_mb": 2.5, "columns_info": [] }
bashuv sync
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 11,565 | 7,404 | -36% | 1 | 1 | 0% | 2,306 | 1,866 | -19% | 0 | 0 | — |
case-02 | fail→pass | 11,798 | 10,141 | -14% | 1 | 1 | 0% | 2,296 | 1,579 | -31% | 0 | 0 | — |
case-03 | fail→fail | 13,616 | 6,880 | -49% | 1 | 1 | 0% | 2,803 | 1,685 | -40% | 0 | 0 | — |
case-04 | pass→pass | 13,775 | 8,230 | -40% | 1 | 1 | 0% | 2,354 | 1,685 | -28% | 0 | 0 | — |
case-05 | pass→pass | 14,877 | 7,639 | -49% | 1 | 1 | 0% | 2,482 | 1,624 | -35% | 0 | 0 | — |
case-06 | pass→pass | 17,860 | 8,811 | -51% | 1 | 1 | 0% | 2,445 | 1,860 | -24% | 0 | 0 | — |
case-07 | fail→pass | 11,485 | 3,435 | -70% | 1 | 1 | 0% | 1,914 | 790 | -59% | 0 | 0 | — |
case-08 | fail→pass | 7,103 | 3,162 | -55% | 1 | 1 | 0% | 1,252 | 816 | -35% | 0 | 0 | — |
case-09 | pass→pass | 11,498 | 5,048 | -56% | 1 | 1 | 0% | 2,023 | 1,036 | -49% | 0 | 0 | — |
case-10 | fail→pass | 14,121 | 4,002 | -72% | 1 | 1 | 0% | 2,175 | 847 | -61% | 0 | 0 | — |
case-11 | pass→pass | 7,060 | 2,863 | -59% | 1 | 1 | 0% | 1,202 | 539 | -55% | 0 | 0 | — |
case-12 | pass→pass | 9,486 | 2,007 | -79% | 1 | 1 | 0% | 1,472 | 478 | -68% | 0 | 0 | — |
case-13 | fail→pass | 8,382 | 2,053 | -76% | 1 | 1 | 0% | 1,500 | 561 | -63% | 0 | 0 | — |
case-14 | pass→pass | 5,748 | 2,847 | -50% | 1 | 1 | 0% | 933 | 668 | -28% | 0 | 0 | — |
case-15 | fail→pass | 8,359 | 2,316 | -72% | 1 | 1 | 0% | 1,456 | 558 | -62% | 0 | 0 | — |
case-16 | pass→pass | 7,308 | 2,766 | -62% | 1 | 1 | 0% | 1,314 | 718 | -45% | 0 | 0 | — |
case-17 | pass→pass | 10,182 | 2,202 | -78% | 1 | 1 | 0% | 1,561 | 605 | -61% | 0 | 0 | — |
case-18 | fail→pass | 11,310 | 3,342 | -70% | 1 | 1 | 0% | 1,847 | 752 | -59% | 0 | 0 | — |
case-19 | fail→pass | 8,095 | 2,423 | -70% | 1 | 1 | 0% | 1,334 | 575 | -57% | 0 | 0 | — |
case-20 | fail→pass | 14,510 | 3,648 | -75% | 1 | 1 | 0% | 2,392 | 828 | -65% | 0 | 0 | — |
case-21 | fail→pass | 11,534 | 3,358 | -71% | 1 | 1 | 0% | 1,877 | 851 | -55% | 0 | 0 | — |
case-22 | fail→pass | 10,116 | 3,504 | -65% | 1 | 1 | 0% | 1,727 | 886 | -49% | 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 +50 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.