Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Windows环境下Claude Code插件Hook执行错误的诊断与修复。当遇到hook error、cannot execute binary file、.sh regex误匹配、WSL/Git Bash冲突时使用。
.claude/skills/foryourhealth111-pixel-windows-hook-debugging/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -17% | 0% |
诊断和修复 Windows 环境下 Claude Code 插件 Hook 执行失败的问题。
触发条件（满足任一即可）：
SessionStart hook error 或 Stop hook errorcannot execute binary file/d/ 或 /c/ 风格的路径（MSYS2/Git Bash 路径格式）bash: node: No such file or directorywsl: 前缀.sh 参数导致 Windows 下被错误处理| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-13 | pass→pass | 8,100 | 8,359 | +3% | 1 | 1 | 0% | 1,574 | 1,693 | +8% | 0 | 0 | — |
case-01 | fail→fail | 15,057 | 13,065 | -13% | 1 | 1 | 0% | 2,395 | 2,460 | +3% | 0 | 0 | — |
case-02 | pass→fail | 11,521 | 9,724 | -16% | 1 | 1 | 0% | 2,275 | 2,057 | -10% | 0 | 0 | — |
case-03 | fail→fail | 11,787 | 9,631 | -18% | 1 | 1 | 0% | 2,091 | 1,932 | -8% | 0 | 0 | — |
case-04 | fail→pass | 11,689 | 9,322 | -20% | 1 | 1 | 0% | 2,034 | 1,799 | -12% | 0 | 0 | — |
case-05 | pass→pass | 8,174 | 9,054 | +11% | 1 | 1 | 0% | 1,469 | 1,772 | +21% | 0 | 0 | — |
case-06 | fail→pass | 11,381 | 8,267 | -27% | 1 | 1 | 0% | 1,913 | 1,709 | -11% | 0 | 0 | — |
case-07 | fail→fail | 11,656 | 13,025 | +12% | 1 | 1 | 0% | 2,253 | 2,513 | +12% | 0 | 0 | — |
case-08 | fail→pass | 11,921 | 8,665 | -27% | 1 | 1 | 0% | 2,080 | 1,690 | -19% | 0 | 0 | — |
case-09 | fail→pass | 13,751 | 9,260 | -33% | 1 | 1 | 0% | 2,459 | 1,834 | -25% | 0 | 0 | — |
case-10 | fail→pass | 10,346 | 7,849 | -24% | 1 | 1 | 0% | 1,849 | 1,542 | -17% | 0 | 0 | — |
case-11 | fail→fail | 11,685 | 9,137 | -22% | 1 | 1 | 0% | 2,035 | 1,806 | -11% | 0 | 0 | — |
case-12 | fail→fail | 10,384 | 10,065 | -3% | 1 | 1 | 0% | 1,822 | 1,988 | +9% | 0 | 0 | — |
case-14 | fail→fail | 10,815 | 6,606 | -39% | 1 | 1 | 0% | 1,789 | 1,247 | -30% | 0 | 0 | — |
case-15 | fail→pass | 10,552 | 7,633 | -28% | 1 | 1 | 0% | 1,837 | 1,531 | -17% | 0 | 0 | — |
case-16 | fail→fail | 9,733 | 9,184 | -6% | 1 | 1 | 0% | 1,769 | 1,918 | +8% | 0 | 0 | — |
case-17 | pass→pass | 8,740 | 8,325 | -5% | 1 | 1 | 0% | 1,618 | 1,796 | +11% | 0 | 0 | — |
case-18 | pass→pass | 8,402 | 6,152 | -27% | 1 | 1 | 0% | 1,516 | 1,281 | -16% | 0 | 0 | — |
case-19 | fail→fail | 11,372 | 6,008 | -47% | 1 | 1 | 0% | 1,858 | 1,216 | -35% | 0 | 0 | — |
case-20 | pass→pass | 11,416 | 6,692 | -41% | 1 | 1 | 0% | 2,152 | 1,424 | -34% | 0 | 0 | — |
case-21 | pass→pass | 14,741 | 9,516 | -35% | 1 | 1 | 0% | 2,725 | 1,759 | -35% | 0 | 0 | — |
case-22 | pass→pass | 13,251 | 11,259 | -15% | 1 | 1 | 0% | 2,471 | 2,368 | -4% | 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 +23 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.