Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Scans directory structure, detects missing documentation, and verifies code-doc synchronization. Use when checking module completeness, README presence, or DESIGN.md alignment. Automatically triggered after creating new modules.
.claude/skills/telagod-verifying-modules/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 4% | 0% |
> 文档不齐不算交付。模块是否能被外人独立理解，是「完成」的最低门槛。
| 场景 | 跑 | 理由 | |------|------|------| | 新建模块 | ✅ 必跑 | 阻断「无文档落地」 | | 模块重构（边界变化） | ✅ | 文档随边界更新 | | 接收他人模块 | ✅ | 检查可理解性 | | 提交前（涉及模块新增） | ✅ | 最后闸 | | 仅函数级修改 | ❌ | 走 analyzing-changes 即可 |
| 文件 | 缺失后果 | 备注 | |------|----------|------| | README.md | 阻断交付 | 用户视角 | | DESIGN.md | 阻断交付 | 维护者视角 | | tests/ | 警告 | 单文件模块或纯类型声明可豁免 | | __init__.py | 提示（仅 Python） | 视项目布局而定 |
_internal/、internal/）→ README 简化，DESIGN 可缺bashnode scripts/module_scanner.js <模块路径> node scripts/module_scanner.js <模块路径> -v # 详细 node scripts/module_scanner.js <模块路径> --json # CI 用
阻断项必修。警告项可在 DESIGN 留「不写测试的理由」（如纯接口声明、纯常量）。「以后补」不算理由。
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 20,751 | 14,529 | -30% | 1 | 1 | 0% | 3,039 | 2,912 | -4% | 0 | 0 | — |
case-02 | fail→pass | 18,337 | 11,131 | -39% | 1 | 1 | 0% | 2,495 | 2,414 | -3% | 0 | 0 | — |
case-03 | fail→pass | 22,834 | 21,397 | -6% | 1 | 1 | 0% | 3,185 | 3,805 | +19% | 0 | 0 | — |
case-04 | fail→pass | 15,543 | 11,447 | -26% | 1 | 1 | 0% | 2,419 | 2,302 | -5% | 0 | 0 | — |
case-05 | fail→fail | 18,325 | 8,635 | -53% | 1 | 1 | 0% | 2,900 | 2,033 | -30% | 0 | 0 | — |
case-06 | fail→pass | 18,530 | 12,555 | -32% | 1 | 1 | 0% | 2,598 | 2,698 | +4% | 0 | 0 | — |
case-07 | fail→pass | 22,776 | 14,260 | -37% | 1 | 1 | 0% | 2,889 | 2,623 | -9% | 0 | 0 | — |
case-08 | pass→pass | 15,035 | 4,938 | -67% | 1 | 1 | 0% | 2,175 | 1,426 | -34% | 0 | 0 | — |
case-09 | fail→pass | 16,222 | 9,793 | -40% | 1 | 1 | 0% | 2,291 | 2,242 | -2% | 0 | 0 | — |
case-10 | fail→pass | 16,040 | 7,779 | -52% | 1 | 1 | 0% | 2,209 | 1,867 | -15% | 0 | 0 | — |
case-11 | pass→pass | 14,490 | 12,939 | -11% | 1 | 1 | 0% | 2,238 | 2,675 | +20% | 0 | 0 | — |
case-12 | pass→pass | 17,639 | 7,894 | -55% | 1 | 1 | 0% | 2,459 | 1,994 | -19% | 0 | 0 | — |
case-13 | pass→pass | 18,642 | 13,422 | -28% | 1 | 1 | 0% | 2,824 | 2,608 | -8% | 0 | 0 | — |
case-14 | pass→pass | 18,183 | 14,523 | -20% | 1 | 1 | 0% | 2,618 | 2,839 | +8% | 0 | 0 | — |
case-15 | pass→pass | 17,062 | 7,809 | -54% | 1 | 1 | 0% | 2,355 | 1,783 | -24% | 0 | 0 | — |
case-16 | pass→pass | 13,968 | 10,998 | -21% | 1 | 1 | 0% | 2,003 | 2,203 | +10% | 0 | 0 | — |
case-17 | pass→pass | 17,482 | 15,555 | -11% | 1 | 1 | 0% | 2,508 | 2,939 | +17% | 0 | 0 | — |
case-18 | pass→pass | 16,242 | 7,336 | -55% | 1 | 1 | 0% | 2,327 | 1,845 | -21% | 0 | 0 | — |
case-19 | fail→pass | 15,623 | 7,788 | -50% | 1 | 1 | 0% | 2,274 | 1,843 | -19% | 0 | 0 | — |
case-20 | pass→pass | 14,955 | 8,997 | -40% | 1 | 1 | 0% | 2,237 | 2,124 | -5% | 0 | 0 | — |
case-21 | pass→pass | 11,063 | 5,665 | -49% | 1 | 1 | 0% | 1,669 | 1,626 | -3% | 0 | 0 | — |
case-22 | pass→pass | 11,224 | 2,444 | -78% | 1 | 1 | 0% | 1,781 | 1,053 | -41% | 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 +41 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.