Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Fetch live cryptocurrency prices with zero API keys — CoinGecko's public endpoints primary, Coinbase spot fallback, via plain curl. Use when asked what's bitcoin at, ETH price in euros, how's the crypto market today, or price of some altcoin. Produces the current price with 24h context, the source and timestamp, the rerunnable command, and the volatility caveat that crypto answers must carry.
.claude/skills/mohitagw15856-crypto-prices/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 62% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 36% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 110% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 32% | 0% |
Crypto prices are the most-asked live number after weather, and two services answer keylessly: CoinGecko's public API (thousands of coins, market context, any quote currency) and Coinbase's spot endpoint (fast, reliable, majors only). This skill fetches, adds the 24-hour context that turns a number into information, and carries the caveat every crypto answer owes: this number is already stale, and it is not advice.
Ask for these if not provided:
https://api.coingecko.com/api/v3/search?query=name — ticker collisions are common and the wrong coin is a real failure modevs_currencies=eur,inr,jpy)curl -s "https://api.coingecko.com/api/v3/simple/price?ids=bitcoin,ethereum&vs_currencies=usd&include_24hr_change=true&include_24hr_vol=true&include_last_updated_at=true" · richer context: .../api/v3/coins/markets?vs_currency=usd&ids=bitcoin,ethereum (adds rank, high/low, ATH distance) · trending: .../api/v3/search/trending. Public tier rate-limits (~10–30 req/min) — batch ids into one call, never loop.curl -s "https://api.coinbase.com/v2/prices/BTC-USD/spot" — extremely reliable, majors and USD/EUR/GBP pairs, no market context. Use when CoinGecko rate-limits or for a fast single number.include_last_updated_at and print it; crypto moves percent-per-hour on bad days, and an undated price is misinformation waiting to age.Price] currency] · 24h: ±x.x%] · rank/context if asked]
Multi-coin questions: small table, one row per coin]
Source: CoinGecko / Coinbase] at timestamp, UTC] · rerun: [exact curl] Live snapshot, already aging — informational only, not investment advice.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-10 | fail→pass | 13,465 | 6,674 | -50% | 1 | 1 | 0% | 1,897 | 2,007 | +6% | 0 | 0 | — |
case-01 | fail→pass | 6,901 | 5,376 | -22% | 1 | 1 | 0% | 1,269 | 2,054 | +62% | 0 | 0 | — |
case-02 | pass→pass | 9,837 | 34,118 | +247% | 1 | 1 | 0% | 1,745 | 2,301 | +32% | 0 | 0 | — |
case-03 | pass→pass | 11,059 | 9,760 | -12% | 1 | 1 | 0% | 2,229 | 2,774 | +24% | 0 | 0 | — |
case-04 | pass→pass | 16,445 | 12,581 | -23% | 1 | 1 | 0% | 3,002 | 3,561 | +19% | 0 | 0 | — |
case-05 | pass→pass | 17,966 | 16,178 | -10% | 1 | 1 | 0% | 3,365 | 3,901 | +16% | 0 | 0 | — |
case-06 | pass→pass | 10,270 | 6,538 | -36% | 1 | 1 | 0% | 1,881 | 2,390 | +27% | 0 | 0 | — |
case-07 | pass→pass | 11,746 | 9,452 | -20% | 1 | 1 | 0% | 2,048 | 2,569 | +25% | 0 | 0 | — |
case-08 | pass→pass | 8,513 | 5,069 | -40% | 1 | 1 | 0% | 1,701 | 1,821 | +7% | 0 | 0 | — |
case-09 | pass→pass | 10,283 | 27,446 | +167% | 1 | 1 | 0% | 1,216 | 1,928 | +59% | 0 | 0 | — |
case-11 | pass→pass | 6,246 | 6,692 | +7% | 1 | 1 | 0% | 996 | 1,953 | +96% | 0 | 0 | — |
case-12 | pass→pass | 7,463 | 6,360 | -15% | 1 | 1 | 0% | 1,213 | 2,125 | +75% | 0 | 0 | — |
case-13 | pass→pass | 10,149 | 5,170 | -49% | 1 | 1 | 0% | 1,533 | 1,905 | +24% | 0 | 0 | — |
case-14 | fail→pass | 9,897 | 6,851 | -31% | 1 | 1 | 0% | 1,538 | 2,096 | +36% | 0 | 0 | — |
case-15 | fail→pass | 5,133 | 3,652 | -29% | 1 | 1 | 0% | 707 | 1,485 | +110% | 0 | 0 | — |
case-16 | pass→pass | 6,378 | 4,968 | -22% | 1 | 1 | 0% | 1,315 | 1,900 | +44% | 0 | 0 | — |
case-17 | pass→pass | 5,735 | 50,160 | +775% | 1 | 1 | 0% | 1,128 | 1,517 | +34% | 0 | 0 | — |
case-18 | pass→pass | 5,402 | 3,932 | -27% | 1 | 1 | 0% | 1,094 | 1,748 | +60% | 0 | 0 | — |
case-19 | pass→pass | 5,326 | 7,484 | +41% | 1 | 1 | 0% | 959 | 2,436 | +154% | 0 | 0 | — |
case-20 | pass→pass | 4,124 | 6,317 | +53% | 1 | 1 | 0% | 661 | 2,089 | +216% | 0 | 0 | — |
case-21 | fail→fail | 2,694 | 8,779 | +226% | 1 | 1 | 0% | 378 | 2,806 | +642% | 0 | 0 | — |
case-22 | pass→pass | 2,514 | 2,641 | +5% | 1 | 1 | 0% | 513 | 1,414 | +176% | 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. 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.