Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Protect and accelerate websites with Cloudflare. Use when a user asks to add CDN, DDoS protection, DNS management, SSL, WAF, or edge computing to a website or API.
.claude/skills/terminalskills-cloudflare/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-01 | ✓→✗ | ▼ Worse | 18% | 0% |
| case-12 | ✓→✓ | = Same ✓ | -38% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 11% | 0% |
| case-03 | ✓→✓ | = Same ✓ | -17% | 0% |
Cloudflare provides CDN, DDoS protection, DNS, SSL, WAF, and edge computing (Workers). Free tier includes unlimited bandwidth, DNS, basic DDoS protection, and SSL.
Point your domain nameservers to Cloudflare, then manage DNS via dashboard or API.
bash# Cloudflare API — manage DNS records curl -X POST "https://api.cloudflare.com/client/v4/zones/ZONE_ID/dns_records" \ -H "Authorization: Bearer CF_API_TOKEN" \ -H "Content-Type: application/json" \ --data '{"type":"A","name":"app","content":"1.2.3.4","proxied":true}'
Always use Full (Strict) mode in production:
hcl# cloudflare.tf — Infrastructure as code resource "cloudflare_record" "app" { zone_id = var.cloudflare_zone_id name = "app" content = "1.2.3.4" type = "A" proxied = true }
javascript// worker.js — Runs at the edge, <1ms cold start export default { async fetch(request) { const url = new URL(request.url) if (url.pathname === '/api/health') { return new Response('OK', { status: 200 }) } return fetch(request) // pass through to origin } }
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-12 | pass→pass | 9,957 | 2,578 | -74% | 1 | 1 | 0% | 1,562 | 961 | -38% | 0 | 0 | — |
case-01 | pass→fail | 6,104 | 5,185 | -15% | 1 | 1 | 0% | 1,217 | 1,431 | +18% | 0 | 0 | — |
case-02 | pass→pass | 7,598 | 6,340 | -17% | 1 | 1 | 0% | 1,605 | 1,779 | +11% | 0 | 0 | — |
case-03 | pass→pass | 6,822 | 3,561 | -48% | 1 | 1 | 0% | 1,352 | 1,128 | -17% | 0 | 0 | — |
case-04 | pass→pass | 8,903 | 3,510 | -61% | 1 | 1 | 0% | 1,847 | 1,130 | -39% | 0 | 0 | — |
case-05 | pass→pass | 3,965 | 2,635 | -34% | 1 | 1 | 0% | 572 | 965 | +69% | 0 | 0 | — |
case-06 | pass→pass | 3,623 | 2,100 | -42% | 1 | 1 | 0% | 589 | 843 | +43% | 0 | 0 | — |
case-07 | pass→pass | 5,162 | 2,562 | -50% | 1 | 1 | 0% | 911 | 976 | +7% | 0 | 0 | — |
case-13 | pass→pass | 3,248 | 1,750 | -46% | 1 | 1 | 0% | 526 | 769 | +46% | 0 | 0 | — |
case-08 | pass→pass | 8,111 | 7,090 | -13% | 1 | 1 | 0% | 1,415 | 1,742 | +23% | 0 | 0 | — |
case-09 | fail→pass | 7,166 | 2,632 | -63% | 1 | 1 | 0% | 1,493 | 964 | -35% | 0 | 0 | — |
case-10 | pass→pass | 5,307 | 2,878 | -46% | 1 | 1 | 0% | 974 | 1,039 | +7% | 0 | 0 | — |
case-11 | pass→pass | 7,937 | 5,693 | -28% | 1 | 1 | 0% | 1,449 | 1,526 | +5% | 0 | 0 | — |
case-14 | pass→pass | 11,357 | 5,658 | -50% | 1 | 1 | 0% | 1,879 | 1,336 | -29% | 0 | 0 | — |
case-15 | pass→pass | 5,743 | 2,898 | -50% | 1 | 1 | 0% | 801 | 957 | +19% | 0 | 0 | — |
case-16 | pass→pass | 7,925 | 2,676 | -66% | 1 | 1 | 0% | 1,246 | 897 | -28% | 0 | 0 | — |
case-17 | pass→pass | 2,371 | 1,509 | -36% | 1 | 1 | 0% | 305 | 698 | +129% | 0 | 0 | — |
case-18 | pass→pass | 6,576 | 3,001 | -54% | 1 | 1 | 0% | 1,204 | 1,008 | -16% | 0 | 0 | — |
case-19 | pass→pass | 2,828 | 4,956 | +75% | 1 | 1 | 0% | 527 | 917 | +74% | 0 | 0 | — |
case-20 | pass→pass | 1,949 | 2,164 | +11% | 1 | 1 | 0% | 317 | 852 | +169% | 0 | 0 | — |
case-21 | pass→pass | 6,292 | 1,816 | -71% | 1 | 1 | 0% | 1,119 | 744 | -34% | 0 | 0 | — |
case-22 | pass→pass | 4,685 | 5,166 | +10% | 1 | 1 | 0% | 747 | 773 | +3% | 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 0 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.