Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Federal Reserve Economic Data (FRED) API client. 800,000+ US and international economic time series: GDP, inflation, unemployment, interest rates, industrial production. Direct data access for macro research.
.claude/skills/mkurman-fredapi/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-17 | ✓→✗ | ▼ Worse | 1% | 0% |
Fred API provides access to Federal Reserve Economic Data (FRED) — 800,000+ US and international economic time series. Use it for GDP, unemployment, inflation, interest rates, industrial production, and financial market data.
bashuv pip install fredapi
pythonfrom fredapi import Fred fred = Fred(api_key="YOUR_API_KEY") # get free key from research.stlouisfed.org # Get GDP data (series ID: GDP) gdp = fred.get_series("GDP") print(gdp.tail()) # Get unemployment rate unemp = fred.get_series("UNRATE") print(unemp.tail())
python# Search for series results = fred.search("consumer price index", limit=5) print(results[["id", "title", "observation_start", "observation_end"]]) # Get series info info = fred.get_series_info("CPIAUCSL") # CPI Urban Consumers print(info)
python# Get multiple series at once cpi = fred.get_series("CPIAUCSL") fed_funds = fred.get_series("FEDFUNDS") ten_year = fred.get_series("DGS10") import pandas as pd combined = pd.DataFrame({"CPI": cpi, "FedFunds": fed_funds, "10Y": ten_year})
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 9,299 | 8,457 | -9% | 1 | 1 | 0% | 1,930 | 2,132 | +10% | 0 | 0 | — |
case-13 | pass→pass | 7,435 | 3,150 | -58% | 1 | 1 | 0% | 1,361 | 928 | -32% | 0 | 0 | — |
case-14 | pass→pass | 10,271 | 4,858 | -53% | 1 | 1 | 0% | 1,896 | 1,237 | -35% | 0 | 0 | — |
case-01 | fail→pass | 6,588 | 3,355 | -49% | 1 | 1 | 0% | 1,261 | 1,067 | -15% | 0 | 0 | — |
case-02 | fail→fail | 11,130 | 7,960 | -28% | 1 | 1 | 0% | 2,036 | 1,988 | -2% | 0 | 0 | — |
case-03 | pass→pass | 10,287 | 6,398 | -38% | 1 | 1 | 0% | 2,152 | 1,507 | -30% | 0 | 0 | — |
case-05 | pass→pass | 12,318 | 12,346 | +0% | 1 | 1 | 0% | 2,330 | 2,913 | +25% | 0 | 0 | — |
case-06 | pass→pass | 4,652 | 3,011 | -35% | 1 | 1 | 0% | 872 | 825 | -5% | 0 | 0 | — |
case-07 | fail→pass | 9,622 | 3,026 | -69% | 1 | 1 | 0% | 1,664 | 896 | -46% | 0 | 0 | — |
case-08 | fail→pass | 7,969 | 2,905 | -64% | 1 | 1 | 0% | 1,458 | 855 | -41% | 0 | 0 | — |
case-09 | pass→pass | 9,130 | 3,534 | -61% | 1 | 1 | 0% | 1,764 | 1,038 | -41% | 0 | 0 | — |
case-10 | pass→pass | 5,454 | 2,229 | -59% | 1 | 1 | 0% | 998 | 759 | -24% | 0 | 0 | — |
case-11 | pass→pass | 5,478 | 2,935 | -46% | 1 | 1 | 0% | 1,052 | 813 | -23% | 0 | 0 | — |
case-12 | pass→pass | 4,046 | 1,435 | -65% | 1 | 1 | 0% | 624 | 536 | -14% | 0 | 0 | — |
case-15 | fail→pass | 5,957 | 1,837 | -69% | 1 | 1 | 0% | 1,015 | 677 | -33% | 0 | 0 | — |
case-16 | pass→pass | 8,867 | 5,029 | -43% | 1 | 1 | 0% | 1,590 | 1,212 | -24% | 0 | 0 | — |
case-17 | pass→fail | 7,402 | 5,643 | -24% | 1 | 1 | 0% | 1,366 | 1,379 | +1% | 0 | 0 | — |
case-18 | pass→pass | 3,869 | 2,335 | -40% | 1 | 1 | 0% | 640 | 771 | +20% | 0 | 0 | — |
case-19 | pass→pass | 2,716 | 2,821 | +4% | 1 | 1 | 0% | 427 | 777 | +82% | 0 | 0 | — |
case-20 | pass→pass | 3,196 | 2,650 | -17% | 1 | 1 | 0% | 425 | 806 | +90% | 0 | 0 | — |
case-21 | pass→pass | 2,669 | 2,050 | -23% | 1 | 1 | 0% | 433 | 679 | +57% | 0 | 0 | — |
case-22 | pass→pass | 3,481 | 3,028 | -13% | 1 | 1 | 0% | 581 | 922 | +59% | 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. 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.