Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Deep HTTP/1.1, HTTP/2, and HTTP/3 protocol expertise for web protocol implementation and compliance
.claude/skills/a5c-ai-http-protocol/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | -49% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -54% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | -18% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 114% | 0% |
Expert skill for HTTP protocol implementation, testing, and compliance validation across HTTP/1.1, HTTP/2, and HTTP/3 versions.
curl - HTTP client with extensive protocol supporth2spec - HTTP/2 conformance testing toolnghttp2 - HTTP/2 library and toolsquiche - HTTP/3 and QUIC implementationhttpie - User-friendly HTTP clientbashh2spec -h localhost -p 8443 --tls --insecure
bashcurl -v -I https://example.com
bashnghttp -v https://example.com
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 12,339 | 8,592 | -30% | 1 | 1 | 0% | 1,450 | 1,193 | -18% | 0 | 0 | — |
case-02 | pass→pass | 8,616 | 10,122 | +17% | 1 | 1 | 0% | 652 | 1,393 | +114% | 0 | 0 | — |
case-03 | pass→pass | 10,190 | 8,419 | -17% | 1 | 1 | 0% | 902 | 1,056 | +17% | 0 | 0 | — |
case-04 | pass→pass | 13,213 | 9,292 | -30% | 1 | 1 | 0% | 1,613 | 1,238 | -23% | 0 | 0 | — |
case-05 | pass→pass | 8,023 | 2,697 | -66% | 1 | 1 | 0% | 546 | 841 | +54% | 0 | 0 | — |
case-06 | pass→pass | 9,435 | 3,790 | -60% | 1 | 1 | 0% | 801 | 1,156 | +44% | 0 | 0 | — |
case-07 | pass→pass | 7,874 | 3,549 | -55% | 1 | 1 | 0% | 1,352 | 990 | -27% | 0 | 0 | — |
case-08 | pass→pass | 3,996 | 3,416 | -15% | 1 | 1 | 0% | 665 | 938 | +41% | 0 | 0 | — |
case-09 | pass→pass | 9,709 | 11,097 | +14% | 1 | 1 | 0% | 869 | 1,314 | +51% | 0 | 0 | — |
case-10 | pass→pass | 7,226 | 1,736 | -76% | 1 | 1 | 0% | 320 | 685 | +114% | 0 | 0 | — |
case-11 | pass→pass | 5,303 | 13,245 | +150% | 1 | 1 | 0% | 936 | 1,756 | +88% | 0 | 0 | — |
case-12 | pass→pass | 8,709 | 3,658 | -58% | 1 | 1 | 0% | 579 | 1,103 | +91% | 0 | 0 | — |
case-13 | fail→pass | 14,367 | 7,454 | -48% | 1 | 1 | 0% | 1,721 | 871 | -49% | 0 | 0 | — |
case-14 | pass→pass | 18,818 | 6,712 | -64% | 1 | 1 | 0% | 2,387 | 712 | -70% | 0 | 0 | — |
case-15 | fail→pass | 9,159 | 1,632 | -82% | 1 | 1 | 0% | 1,499 | 692 | -54% | 0 | 0 | — |
case-16 | fail→pass | 6,291 | 6,589 | +5% | 1 | 1 | 0% | 1,133 | 670 | -41% | 0 | 0 | — |
case-17 | pass→pass | 7,480 | 8,634 | +15% | 1 | 1 | 0% | 402 | 822 | +104% | 0 | 0 | — |
case-18 | pass→pass | 8,368 | 8,891 | +6% | 1 | 1 | 0% | 600 | 1,120 | +87% | 0 | 0 | — |
case-19 | pass→pass | 8,040 | 7,793 | -3% | 1 | 1 | 0% | 513 | 912 | +78% | 0 | 0 | — |
case-20 | pass→pass | 20,596 | 21,299 | +3% | 1 | 1 | 0% | 3,040 | 3,684 | +21% | 0 | 0 | — |
case-21 | pass→pass | 19,430 | 15,009 | -23% | 1 | 1 | 0% | 2,922 | 2,407 | -18% | 0 | 0 | — |
case-22 | pass→pass | 10,205 | 9,472 | -7% | 1 | 1 | 0% | 1,902 | 2,133 | +12% | 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 +14 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.