Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing templates, rendered HTML, or shared components related to Check for broken links. Validate the final browser-facing markup, not just the source framework abstraction.
.claude/skills/thedaviddias-link-checker/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 110% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 225% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-07 | ✓→✓ | ＝ Same ✓ | -9% | 0% |
Broken links frustrate users, hurt SEO rankings (Google penalizes 404s), and damage credibility—especially for documentation and e-commerce sites.
Verify that all internal and external links are working properly and don't return 404 errors or redirect to unintended destinations.
Use automated link checking tools to identify and fix broken links, update redirected URLs, and remove or replace dead links.
Explain why broken links hurt user experience, SEO rankings, and credibility, and how to implement ongoing link monitoring.
Review templates, server-rendered HTML, and shared components that output markup related to Check for broken links. Flag exact elements, attributes, and routes where the rendered HTML violates the rule.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/html/link-checker
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-07 | pass→pass | 16,051 | 12,121 | -24% | 1 | 1 | 0% | 2,185 | 1,987 | -9% | 0 | 0 | — |
case-01 | fail→pass | 9,122 | 17,868 | +96% | 1 | 1 | 0% | 1,393 | 2,928 | +110% | 0 | 0 | — |
case-02 | fail→pass | 13,030 | 15,978 | +23% | 1 | 1 | 0% | 2,101 | 2,551 | +21% | 0 | 0 | — |
case-03 | fail→fail | 15,853 | 14,149 | -11% | 1 | 1 | 0% | 2,506 | 2,599 | +4% | 0 | 0 | — |
case-04 | pass→pass | 14,900 | 14,663 | -2% | 1 | 1 | 0% | 2,672 | 2,630 | -2% | 0 | 0 | — |
case-05 | fail→pass | 3,694 | 7,482 | +103% | 1 | 1 | 0% | 454 | 1,476 | +225% | 0 | 0 | — |
case-06 | pass→pass | 16,365 | 14,054 | -14% | 1 | 1 | 0% | 2,513 | 2,504 | -0% | 0 | 0 | — |
case-08 | pass→pass | 17,193 | 11,409 | -34% | 1 | 1 | 0% | 2,384 | 2,124 | -11% | 0 | 0 | — |
case-09 | fail→pass | 16,213 | 12,873 | -21% | 1 | 1 | 0% | 2,440 | 2,180 | -11% | 0 | 0 | — |
case-10 | pass→pass | 15,313 | 12,535 | -18% | 1 | 1 | 0% | 1,989 | 1,993 | +0% | 0 | 0 | — |
case-11 | pass→pass | 17,972 | 14,495 | -19% | 1 | 1 | 0% | 2,371 | 2,416 | +2% | 0 | 0 | — |
case-12 | pass→pass | 16,283 | 15,216 | -7% | 1 | 1 | 0% | 2,232 | 2,390 | +7% | 0 | 0 | — |
case-13 | pass→pass | 17,730 | 9,791 | -45% | 1 | 1 | 0% | 2,547 | 2,213 | -13% | 0 | 0 | — |
case-14 | pass→pass | 13,568 | 11,981 | -12% | 1 | 1 | 0% | 1,751 | 1,885 | +8% | 0 | 0 | — |
case-15 | pass→pass | 13,034 | 10,036 | -23% | 1 | 1 | 0% | 1,826 | 1,623 | -11% | 0 | 0 | — |
case-16 | pass→pass | 13,490 | 6,586 | -51% | 1 | 1 | 0% | 1,955 | 1,161 | -41% | 0 | 0 | — |
case-17 | pass→pass | 11,544 | 7,794 | -32% | 1 | 1 | 0% | 1,740 | 1,350 | -22% | 0 | 0 | — |
case-18 | pass→pass | 14,627 | 6,880 | -53% | 1 | 1 | 0% | 2,079 | 1,260 | -39% | 0 | 0 | — |
case-19 | fail→fail | 11,508 | 10,278 | -11% | 1 | 1 | 0% | 1,801 | 1,719 | -5% | 0 | 0 | — |
case-20 | pass→pass | 14,407 | 9,067 | -37% | 1 | 1 | 0% | 2,127 | 1,649 | -22% | 0 | 0 | — |
case-21 | pass→pass | 13,950 | 8,964 | -36% | 1 | 1 | 0% | 2,183 | 1,735 | -21% | 0 | 0 | — |
case-22 | pass→pass | 20,083 | 9,771 | -51% | 1 | 1 | 0% | 3,096 | 1,967 | -36% | 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.
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.