worldbank-data360

CLI and TypeScript SDK for the World Bank Data360 API. Search 12,000+ indicators across 100+ databases — GDP, population, climate, health, education. No API key required.

Install

npm install -g worldbank-data360

Or run without installing:

npx worldbank-data360 discover

Quick start

worldbank discover

Prints a guide with examples. Follow the steps — search, check coverage, fetch.

CLI

Search

worldbank search "life expectancy" --top 5 --database WB_WDI

{
  "total": 549,
  "shown": 5,
  "items": [
    {
      "id": "WB_WDI_SP_DYN_LE00_IN",
      "name": "Life expectancy at birth, total (years)",
      "databaseId": "WB_WDI",
      "score": 76.2,
      "topics": ["People", "Health"]
    }
  ]
}

The indicator ID encodes the database: WB_WDI_SP_DYN_LE00_IN → database WB_WDI.

Check coverage

Before fetching, see which countries and years actually have data:

worldbank info WB_WDI_SP_DYN_LE00_IN

{
  "name": "Life expectancy at birth, total (years)",
  "database": "World Development Indicators (WDI)",
  "areas": 265,
  "years": "1960–2024"
}

If years is a list, the indicator is not published annually — gaps are expected. info also shows disaggregation dimensions (sex, age, urbanisation) when available.

Fetch data

worldbank data WB_WDI --indicator WB_WDI_SP_DYN_LE00_IN --area POL,DEU,USA --from 2010 --to 2023

Multiple countries → records grouped by country code automatically:

{
  "count": 42,
  "indicator": "WB_WDI_SP_DYN_LE00_IN",
  "indicatorName": "Life expectancy at birth, total (years)",
  "database": "WB_WDI",
  "meta": { "UNIT_MEASURE": "YR" },
  "records": {
    "DEU": [{ "period": "2010", "value": 80.5 }, ...],
    "POL": [...],
    "USA": [...]
  }
}

Single country → flat array. Repeated metadata in meta, not in every record. SDMX placeholders (_T, _Z) stripped automatically. Pagination is automatic — API caps at 1000 records per call, SDK fetches all pages.

Indicators with sex/age breakdowns show dimensions per record:

worldbank data WB_HCI --indicator WB_HCI_HCI --area FIN,SWE --from 2018 --to 2020

{
  "records": {
    "FIN": [
      { "period": "2018", "value": 0.814 },
      { "period": "2018", "SEX": "M", "value": 0.786 },
      { "period": "2018", "SEX": "F", "value": 0.844 }
    ]
  }
}

Export to CSV

Open directly in Excel or Google Sheets:

worldbank data WB_WDI --indicator WB_WDI_NY_GDP_PCAP_CD --area POL,DEU,USA --from 2010 --to 2023 --format csv > gdp.csv

area,period,value
DEU,2010,40408.71
POL,2010,12602.34
USA,2010,48466.73
...

Sex/age breakdowns become columns automatically.

All commands

Command	Description
worldbank discover	Quick start guide with examples
worldbank discover --databases	List all 100+ databases as JSON
worldbank search <query> [--top N] [--database ID]	Search indicators by keyword
worldbank info <INDICATOR_ID>	Show name, countries, year range, dimensions
worldbank data <DB> --indicator <ID> [--area] [--from] [--to] [--top N] [--all] [--format json|csv]	Fetch data
worldbank explain <DB> --indicator <ID> [...]	Preview query without fetching
worldbank countries	List all countries with ISO alpha-3 codes

Hints appear in stderr after each command — pipe 2>/dev/null to silence.

SDK

npm install worldbank-data360

import { WorldBankClient } from 'worldbank-data360'

const client = new WorldBankClient()

const result = await client
  .data('WB_WDI')
  .indicator('WB_WDI_NY_GDP_PCAP_CD')
  .area(['POL', 'DEU', 'USA'])
  .from('2010').to('2023')
  .fetch()

Data queries

const result = await client
  .data('WB_WDI')
  .indicator('WB_WDI_SP_POP_TOTL')
  .area(['POL', 'DEU', 'USA'])
  .from('2000').to('2023')
  .sex('F')           // demographic filters
  .age('Y15T24')
  .urbanisation('U')
  .fetch()

Search

const results = await client
  .search()
  .search('poverty')
  .database('WB_WDI')
  .top(10)
  .fetchItems()
// → { total, shown, items: [{ id, name, databaseId, score, topics? }] }

Discover databases

const overview = await client.discover()
// → { totalIndicators: 12938, databases: [{ id: 'WB_WDI', name: '...', indicatorCount: 1534 }, ...] }

AI-first methods

// Preview query without fetching
const info = client.data('WB_WDI').indicator('WB_WDI_SP_POP_TOTL').area(['POL']).from('2010').to('2023').explain()

// Format as markdown table for an LLM prompt
const ctx = await client.data('WB_WDI').indicator('WB_WDI_SP_POP_TOTL').area('POL').from('2020').to('2023').toContext()

Error handling

import { SDKRequestError } from 'worldbank-data360'

try {
  await client.data('INVALID_DB').fetch()
} catch (err) {
  if (err instanceof SDKRequestError) {
    console.log(err.sdkError.message)    // "HTTP 400 from GET /data360/data"
    console.log(err.sdkError.suggestion) // "Check your DATABASE_ID..."
  }
}

API reference

data360.worldbank.org/en/api

Method	Endpoint
.discover()	POST /data360/searchv2 (facets)
.data()	GET /data360/data
.search()	POST /data360/searchv2
.disaggregation()	GET /data360/disaggregation
.countries()	GET /V2/country (World Bank V2)

Both APIs are public — no authentication required.

Data attribution

Data sourced from the World Bank Data360 API and World Bank V2 API. Subject to World Bank Terms of Use.

Unofficial open-source wrapper, not affiliated with or endorsed by the World Bank Group.

License

MIT