Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →AI エージェントが大量の実装出力を生成するチームのためのエンジニアリング運用モデル。
.claude/skills/affaan-m-ai-first-engineering/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-07 | ✓→✗ | ▼ Worse | -27% | 0% |
| case-17 | ✓→✗ | ▼ Worse | -1% | 0% |
在为由人工智能辅助代码生成的团队设计流程、评审和架构时，使用此技能。
优先选择对智能体友好的架构：
避免隐含的行为分散在隐藏的惯例中。
评审关注：
尽量减少花在已由自动化覆盖的风格问题上的时间。
强大的人工智能优先工程师：
提高生成代码的测试标准：
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 17,916 | 9,715 | -46% | 1 | 1 | 0% | 2,466 | 1,697 | -31% | 0 | 0 | — |
case-05 | pass→pass | 17,199 | 8,062 | -53% | 1 | 1 | 0% | 2,388 | 1,432 | -40% | 0 | 0 | — |
case-06 | pass→pass | 16,021 | 10,441 | -35% | 1 | 1 | 0% | 2,564 | 1,822 | -29% | 0 | 0 | — |
case-01 | fail→pass | 15,875 | 15,994 | +1% | 1 | 1 | 0% | 2,437 | 2,391 | -2% | 0 | 0 | — |
case-02 | fail→pass | 8,399 | 8,455 | +1% | 1 | 1 | 0% | 1,308 | 1,059 | -19% | 0 | 0 | — |
case-03 | pass→pass | 10,514 | 4,963 | -53% | 1 | 1 | 0% | 1,674 | 1,097 | -34% | 0 | 0 | — |
case-07 | pass→fail | 7,643 | 4,034 | -47% | 1 | 1 | 0% | 1,252 | 916 | -27% | 0 | 0 | — |
case-08 | pass→pass | 10,610 | 4,169 | -61% | 1 | 1 | 0% | 1,647 | 1,008 | -39% | 0 | 0 | — |
case-09 | pass→pass | 9,302 | 5,872 | -37% | 1 | 1 | 0% | 1,497 | 1,150 | -23% | 0 | 0 | — |
case-10 | pass→pass | 8,912 | 7,303 | -18% | 1 | 1 | 0% | 1,497 | 1,382 | -8% | 0 | 0 | — |
case-11 | pass→pass | 10,773 | 5,400 | -50% | 1 | 1 | 0% | 1,604 | 1,126 | -30% | 0 | 0 | — |
case-12 | pass→pass | 12,666 | 10,622 | -16% | 1 | 1 | 0% | 2,046 | 2,081 | +2% | 0 | 0 | — |
case-13 | fail→fail | 9,285 | 3,919 | -58% | 1 | 1 | 0% | 1,394 | 878 | -37% | 0 | 0 | — |
case-14 | pass→pass | 27,045 | 10,942 | -60% | 1 | 1 | 0% | 1,802 | 1,781 | -1% | 0 | 0 | — |
case-15 | fail→pass | 12,394 | 8,761 | -29% | 1 | 1 | 0% | 1,836 | 1,593 | -13% | 0 | 0 | — |
case-16 | pass→pass | 10,676 | 7,263 | -32% | 1 | 1 | 0% | 1,517 | 1,294 | -15% | 0 | 0 | — |
case-17 | pass→fail | 15,836 | 14,507 | -8% | 1 | 1 | 0% | 2,431 | 2,396 | -1% | 0 | 0 | — |
case-18 | pass→pass | 14,005 | 11,250 | -20% | 1 | 1 | 0% | 2,063 | 2,000 | -3% | 0 | 0 | — |
case-19 | pass→pass | 9,449 | 4,442 | -53% | 1 | 1 | 0% | 1,442 | 929 | -36% | 0 | 0 | — |
case-20 | pass→pass | 10,922 | 9,472 | -13% | 1 | 1 | 0% | 1,648 | 1,726 | +5% | 0 | 0 | — |
case-21 | pass→pass | 8,233 | 3,856 | -53% | 1 | 1 | 0% | 1,180 | 761 | -36% | 0 | 0 | — |
case-22 | pass→pass | 9,604 | 7,912 | -18% | 1 | 1 | 0% | 1,691 | 1,568 | -7% | 0 | 0 | — |
case-23 | pass→pass | 8,986 | 6,858 | -24% | 1 | 1 | 0% | 1,601 | 1,396 | -13% | 0 | 0 | — |
case-24 | pass→pass | 12,224 | 10,781 | -12% | 1 | 1 | 0% | 2,326 | 2,190 | -6% | 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. 24 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 24 comparable cases. 2 cases got worse with the skill loaded, and they are 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.