Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Background context for the Electronic Bookkeeping Act (電子帳簿保存法) in the shinkoku tax filing plugin. Contains requirements for electronic bookkeeping, scanner storage, mandatory electronic transaction data storage, and shinkoku's compliance status. This skill is not user-invocable — Claude loads it automatically when responding to electronic bookkeeping compliance questions.
.claude/skills/kazukinagata-tax-ebookkeeping-context/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 1% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -48% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -9% | 0% |
このスキルは電子帳簿保存法の要件・対応状況に関するコンテキストを提供する。
電子帳簿保存法に関する回答を行う際は、references/electronic-bookkeeping.md を読み込んで以下を実行する:
| ファイル | 内容 | |---------|------| | references/electronic-bookkeeping.md | 電帳法3カテゴリ・要件・猶予措置・shinkoku対応状況・令和9年改正対応 |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 19,500 | 10,641 | -45% | 1 | 1 | 0% | 3,021 | 2,091 | -31% | 0 | 0 | — |
case-02 | fail→fail | 17,829 | 11,624 | -35% | 1 | 1 | 0% | 2,990 | 2,264 | -24% | 0 | 0 | — |
case-03 | fail→pass | 29,951 | 13,482 | -55% | 1 | 1 | 0% | 2,792 | 2,423 | -13% | 0 | 0 | — |
case-04 | pass→pass | 14,920 | 11,251 | -25% | 1 | 1 | 0% | 2,508 | 2,097 | -16% | 0 | 0 | — |
case-05 | pass→pass | 15,552 | 13,196 | -15% | 1 | 1 | 0% | 2,435 | 2,404 | -1% | 0 | 0 | — |
case-06 | fail→pass | 22,430 | 11,373 | -49% | 1 | 1 | 0% | 2,247 | 2,262 | +1% | 0 | 0 | — |
case-07 | fail→pass | 17,072 | 6,835 | -60% | 1 | 1 | 0% | 2,669 | 1,387 | -48% | 0 | 0 | — |
case-08 | pass→pass | 16,094 | 11,432 | -29% | 1 | 1 | 0% | 2,496 | 2,083 | -17% | 0 | 0 | — |
case-09 | pass→pass | 18,615 | 11,688 | -37% | 1 | 1 | 0% | 2,955 | 2,312 | -22% | 0 | 0 | — |
case-10 | pass→pass | 8,809 | 6,861 | -22% | 1 | 1 | 0% | 1,512 | 1,365 | -10% | 0 | 0 | — |
case-11 | pass→pass | 10,318 | 9,949 | -4% | 1 | 1 | 0% | 1,736 | 1,876 | +8% | 0 | 0 | — |
case-12 | pass→pass | 15,597 | 6,598 | -58% | 1 | 1 | 0% | 2,297 | 1,293 | -44% | 0 | 0 | — |
case-13 | pass→pass | 12,734 | 9,262 | -27% | 1 | 1 | 0% | 2,236 | 1,606 | -28% | 0 | 0 | — |
case-14 | fail→pass | 15,019 | 11,793 | -21% | 1 | 1 | 0% | 2,592 | 2,346 | -9% | 0 | 0 | — |
case-15 | pass→pass | 15,408 | 10,158 | -34% | 1 | 1 | 0% | 2,495 | 2,010 | -19% | 0 | 0 | — |
case-16 | pass→pass | 6,028 | 4,228 | -30% | 1 | 1 | 0% | 1,094 | 993 | -9% | 0 | 0 | — |
case-17 | pass→pass | 24,214 | 9,821 | -59% | 1 | 1 | 0% | 2,150 | 2,021 | -6% | 0 | 0 | — |
case-18 | fail→fail | 10,215 | 8,136 | -20% | 1 | 1 | 0% | 1,670 | 1,696 | +2% | 0 | 0 | — |
case-19 | pass→pass | 4,334 | 1,990 | -54% | 1 | 1 | 0% | 740 | 510 | -31% | 0 | 0 | — |
case-20 | pass→fail | 13,137 | 5,791 | -56% | 1 | 1 | 0% | 2,322 | 1,226 | -47% | 0 | 0 | — |
case-21 | pass→pass | 7,828 | 5,495 | -30% | 1 | 1 | 0% | 1,607 | 1,295 | -19% | 0 | 0 | — |
case-22 | pass→pass | 7,519 | 4,066 | -46% | 1 | 1 | 0% | 1,129 | 896 | -21% | 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 +18 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
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.