# Lilambd Data for Agents: full reference

> 40 APIs. Pay per call with x402 v2 (USDC, eip155:8453); add sample=1 for a free 3-row preview. Only 200 responses are charged. MCP: https://api.lilambd.com/mcp. Contact: hello@lilambd.com.

## AI Agent Economy Census

- id / MCP tool: agents
- base: https://api.lilambd.com/agents
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/agents/mcp?sample=1
- data as of: 2026-11-03; refresh: x402 every 5 minutes until a day is complete, MCP every 20 minutes, HF and counters daily; product view refreshed hourly; licence: Facts measured by Lilambd; listing texts belong to their publishers
- coverage: about 92,700 mainnet x402 endpoints on 3,260 hosts; 50,000 MCP Registry servers; HF top 100 models/datasets/Spaces; npm/GitHub counters
- sources: x402 Bazaar discovery API (Coinbase CDP) (https://docs.cdp.coinbase.com/x402); Official MCP Registry (https://registry.modelcontextprotocol.io); Hugging Face Hub (https://huggingface.co); npm registry (https://www.npmjs.com); GitHub search (https://github.com)
- known limits: History starts 2026-11-02. Categories come from keywords and can be wrong; about 60% of endpoints fall in other. MCP first_seen is the census day or the version publish date, not a true birth date for servers seen on the first day. Prices in USD only for USDC on Base, Base Sepolia and Solana.

What AI agents can buy, and at what price, measured every day.

- **x402 paid endpoints**: every resource listed in the x402 Bazaar discovery API (about 92,700 on mainnet, 3,260 hosts), with price in USD, network, category, first-seen and last-seen day
- **Price check**: give a price per call and a category, get the share of competing endpoints that are cheaper or dearer plus the price quantiles. Built for anyone pricing an agent-facing API
- **Categories and hosts**: endpoints, hosts and median / p90 price per category; biggest hosts with their main category
- **New today**: change feed of endpoints and MCP servers that appeared since a given day
- **MCP servers**: 50,000 servers from the official MCP Registry with remotes, packages and version dates; search, namespaces, daily publish counts
- **Trending**: Hugging Face top models, datasets and Spaces and top npm packages for x402 / MCP, kept day by day
- **Counters**: daily totals of GitHub topics and npm keywords (mcp, x402, ai-agents, llm, a2a...)

History starts 2026-11-02 and grows by one day every day. Categories are keyword-based and labelled as such. Sources: x402 Bazaar discovery API (Coinbase CDP), registry.modelcontextprotocol.io, Hugging Face Hub, npm registry, GitHub search. Listing texts belong to their publishers.

Routes:
- GET /agents/mcp: Search MCP servers Params: q (Text in name, title or description), remote (true = only servers with a remote URL), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /agents/x402: Search paid endpoints Params: q (Text in description or URL), category (Category (see /x402/categories)), host (Exact host), min_price (Minimum USD per call), max_price (Maximum USD per call), sort (newest | price | cheapest), include_testnet (true to include testnets), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /agents/mcp/new: Newly published MCP servers Params: since (YYYY-MM-DD), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /agents/summary: Today's census Params: include_testnet (true to include testnets)
- GET /agents/counters: Ecosystem counters Params: key (e.g. npm keywords:mcp), days (Days back)
- GET /agents/x402/new: New endpoints Params: since (YYYY-MM-DD, default yesterday), category (Category), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /agents/mcp/daily: MCP servers per day Params: days (How many days back (max 400))
- GET /agents/mcp/{name}: Get an MCP server Params: name*
- GET /agents/x402/daily: Daily new endpoints
- GET /agents/x402/hosts: Hosts Params: category (Category), q (Host text), sort (endpoints | price), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /agents/mcp/namespaces: MCP namespaces Params: limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /agents/trending/{kind}: Trending lists Params: kind*, day (YYYY-MM-DD, default latest), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /agents/x402/categories: Prices by category
- GET /agents/x402/price-check: Check a price Params: price* (USD per call), category (Category, empty = all)

## Airports of the World

- id / MCP tool: airports
- base: https://api.lilambd.com/airports
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/airports/nearby?lat=35.68&lon=139.76&sample=1
- data as of: 2026-11-01; refresh: static snapshot (source changes daily; refresh planned); licence: Public domain
- coverage: snapshot of the community-maintained list
- sources: OurAirports (https://ourairports.com/data/)
- known limits: Small airfields and heliports include duplicates marked "[Duplicate]" by the source.

Airport lookup for travel apps, flight tools, logistics, games and AI agents.

- **86,329 places**: 1,176 large and ~4,500 medium airports plus small airfields, heliports and seaplane bases
- IATA, ICAO, GPS and local codes; coordinates; elevation; ISO country and region; city
- **Nearby** airports for any coordinate with distance in km
- Filter by country, type, scheduled service

Source: OurAirports (public domain).

Routes:
- GET /airports/nearby: Nearby airports Params: lat* (Latitude), lon* (Longitude), type (large_airport | medium_airport | small_airport | heliport), limit (How many (max 50))
- GET /airports/search: Search airports Params: q (Name or city), country (ISO 3166-1 alpha-2), type (Airport type), scheduled (true | false), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /airports/{code}: Get an airport Params: code*

## Biodiversity by Place

- id / MCP tool: biodiversity
- base: https://api.lilambd.com/biodiversity
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/biodiversity/gaps?sample=1
- data as of: 2026-10-01; refresh: one-off snapshot (GBIF 2026-10-01); live GBIF species lookup in /place; licence: Counts only from CC0 1.0 and CC BY 4.0 GBIF records (CC BY 4.0 attribution: cite GBIF.org download doi 10.15468/dl.tup2g6); CC BY-NC records excluded
- coverage: World, 0.5 degree grid: 214,398 cells (91,815 land), 180,857 with records; 231 countries; 3,398 admin1 regions; 3,026,672,393 open-licence occurrence records
- sources: GBIF.org (1 October 2026) GBIF Occurrence Download, CC0 and CC BY 4.0 records (https://doi.org/10.15468/dl.tup2g6); Lilambd world-biodiversity-cells (Hugging Face) (https://huggingface.co/datasets/Lilambd/world-biodiversity-cells); GBIF API (live species facets, GBIF Backbone Taxonomy) (https://api.gbif.org/v1/)
- known limits: Counts are recording effort, not abundance. Species richness per cell counts distinct GBIF species keys with open-licence records. No per-cell species list in the snapshot: top species in /place are live from api.gbif.org (6 s timeout, may be null). Admin1 ids are Lilambd world mesh ids (e.g. jpn_1860), not ISO 3166-2. Per-species lookup and class tables were not loaded. GBIF issue flags were not applied, so some mis-located records remain.

How well is life recorded here? One API for any point on Earth, built from the GBIF occurrence snapshot of 1 October 2026 and counted on a fixed 0.5 degree grid (about 55 km).

Only records licensed **CC0 or CC BY 4.0** are counted (3,026,672,393 records, 180,857 recorded cells); CC BY-NC records are left out so the data can be used commercially with attribution.

- **Place** (finished answer): give lat/lon, get the cell's records and species, decade history, kingdom and record-type mix, world and national percentile, a plain verdict (e.g. "among the least recorded 20% of land cells"), the 24 surrounding cells within 1 degree, admin1 and country context, plus the top recorded species there (live from GBIF, open licences only)
- **Cells**: 214,398 cells (91,815 land cells) by id, by coordinate, by country or admin1 region, sorted by species or records; **nearby** cells in rings around a point
- **Countries**: 231 countries ranked by records, records per 1,000 km2, richest cell, share of 2020s records or share of land cells without any record; full country profile with richest and least-recorded cells
- **Admin1**: 3,398 first-level regions (Lilambd world mesh ids such as jpn_1860) with records per area and empty cells
- **Gaps**: the least-recorded land cells, regions and countries first; `stale=true` finds cells with records but nothing since 2010
- **Decades**: records per decade for the world or one country

Counts follow where people look (birds are most records; cities and birding hubs dominate), so compare a cell with its neighbours and its own past. Cell ids: `cell_id = row*720+col` with row from 90N; `cell_0p5 = floor((lat+90)/0.5)*720+floor((lon+180)/0.5)` is also returned and accepted.

Source: GBIF.org (1 October 2026) GBIF Occurrence Download https://doi.org/10.15468/dl.tup2g6 (CC0 and CC BY 4.0 records), aggregated by Lilambd (Hugging Face Lilambd/world-biodiversity-cells); live species lists from api.gbif.org (CC0 and CC BY 4.0 records), names from the GBIF Backbone Taxonomy (CC BY 4.0).

Routes:
- GET /biodiversity/gaps: Least-recorded places Params: level (cells | admin1 | countries), country (ISO 3166-1 alpha-2), admin1 (Admin1 code), max_records (Cells: at most this many records (default 100)), stale (Cells: true = recorded before but nothing since 2010), min_cells (Admin1/countries: minimum land cells), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /biodiversity/cells: List cells Params: country (ISO 3166-1 alpha-2), admin1 (Admin1 id as returned by /admin1 (e.g. jpn_1860)), lat (Latitude (returns one cell)), lon (Longitude), min_records (Minimum records), max_records (Maximum records), sort (species | records | records_asc | recent), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /biodiversity/place: Biodiversity at a place Params: lat* (Latitude), lon* (Longitude), top (Top species to return (1-25)), species (false to skip the live GBIF species lookup)
- GET /biodiversity/admin1: Admin1 regions of a country Params: country* (ISO 3166-1 alpha-2), by (records | per_km2 | gaps), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /biodiversity/nearby: Cells around a point Params: lat* (Latitude), lon* (Longitude), radius_cells (Rings (1-6)), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /biodiversity/decades: Records by decade Params: country (ISO 3166-1 alpha-2)
- GET /biodiversity/summary: World summary
- GET /biodiversity/countries: Rank countries Params: by (records | per_km2 | species | recent | no_record_share), continent (Continent text), min_cells (Minimum land cells), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /biodiversity/admin1/{code}: Admin1 region Params: code* (Admin1 id from /admin1)
- GET /biodiversity/cells/{cell_id}: Get a cell Params: cell_id* (cell_id (row*720+col from 90N), or cell_0p5 with scheme=cell_0p5), scheme (cell_id (default) | cell_0p5)
- GET /biodiversity/countries/{iso2}: Country profile Params: iso2* (ISO 3166-1 alpha-2 or alpha-3)

## World Calendars and Date Conversion API

- id / MCP tool: calendars
- base: https://api.lilambd.com/calendars
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/calendars/days?sample=1
- data as of: 2026-11-03; refresh: yearly (November, next 5 years) via HF Lilambd/world-calendars; licence: Unicode License v3 (CLDR); computed data by Lilambd; no holidays (koyomi)
- coverage: 15 calendar systems; day table 2026-01-01..2030-12-31 in 11 calendars (1,826 days); arithmetic Julian/Coptic/Ethiopic/Buddhist/ROC/Japanese era/JDN for any date; weekday names 18,729 and era names 59,130 in 657 locales; week rules for 250 countries; 212 festival dates; 120 solar terms
- sources: Unicode CLDR 49 (cldr-json) (https://github.com/unicode-org/cldr-json); Lilambd world-calendars (computed with hijridate, convertdate, lunardate, PyMeeus) (https://huggingface.co/datasets/Lilambd/world-calendars); Reingold & Dershowitz, Calendrical Calculations 4th ed. (rules) (https://www.cambridge.org/9781107057623)
- known limits: Hijri follows Umm al-Qura (sighting countries may differ 1-2 days); Hijri/Hebrew/Persian/Chinese/Saka only 2026-2030; month names not loaded yet; rule texts English in leap_rule_en, original columns Japanese

Convert dates between the world's calendars and get weekday and era names in 657 locales.

- **15 calendar systems** (Gregorian, Julian, Hijri Umm al-Qura / civil / sighted, Hebrew, Chinese, Korean Dangi, Persian, Coptic, Ethiopic, Indian Saka, Buddhist, Japanese, ROC) with epoch, months, year start, day start, leap rule and the countries that use them
- **Date conversion (finished answer)**: one Gregorian date to JDN, Julian, Coptic, Ethiopic, Buddhist, ROC and Japanese era for any date, plus Hijri, Hebrew, Persian, Chinese (with leap month and day ganzhi) and Saka for all 1,826 days of 2026-2030, with weekday name, festivals and solar term on that day
- **Reverse conversion** from Julian, Coptic, Ethiopic, Buddhist, ROC, Hijri, Hebrew, Persian, Chinese or Saka to Gregorian
- **Localised names** from Unicode CLDR 49: 18,729 weekday names and 59,130 era names across 657 locales (month names follow in the next update)
- **Week rules** for 250 countries: first day of week, weekend (232 Sat-Sun, 14 Fri-Sat), preferred calendars
- **212 festival dates** 2026-2030 computed from their own calendars (Easter, Ramadan, Eid, Rosh Hashanah, Passover, Lunar New Year, Mid-Autumn, Nowruz...) and **120 solar terms** to the minute

No holidays, sunrise or moon data here. Sources: Unicode CLDR 49, Reingold & Dershowitz rules, computed by Lilambd.

Routes:
- GET /calendars/days: Rows of the day table: each day 2026-2030 in 11 calendars Params: from (First date YYYY-MM-DD), to (Last date YYYY-MM-DD), limit (Rows per page (1-200, default 50)), offset (Rows to skip for paging (default 0))
- GET /calendars/convert: Finished answer: one Gregorian date in every calendar, with weekday name, festivals and solar term that day Params: date (Gregorian date YYYY-MM-DD (default today)), locale (CLDR locale for the weekday name)
- GET /calendars/systems: All 15 calendar systems with English rules, epoch and how many countries use them Params: locale (Not used here; kept for symmetry)
- GET /calendars/era-names: Era names for a calendar in a locale (CLDR 49) Params: calendar (CLDR calendar id), locale (CLDR locale), width (wide, abbreviated or narrow), limit (Rows per page (1-200, default 50)), offset (Rows to skip for paging (default 0))
- GET /calendars/festivals: Festival dates 2026-2030 computed from their own calendars, with the countries that have a public holiday that day Params: year (Year 2026-2030), tradition (Islam, Judaism, Christian, Chinese...), festival_id (Festival id), country (ISO alpha-2: only festivals that are a holiday there), limit (Rows per page (1-200, default 50)), offset (Rows to skip for paging (default 0))
- GET /calendars/solar-terms: The 24 solar terms with UTC time to the second and the date in Beijing and Tokyo Params: year (Year 2026-2030), limit (Rows per page (1-200, default 50)), offset (Rows to skip for paging (default 0))
- GET /calendars/systems/{id}: One calendar system with its eras Params: id* (calendar_id from /systems), locale (CLDR locale (default en))
- GET /calendars/to-gregorian: Convert a date in another calendar to Gregorian (arithmetic calendars any year; table calendars 2026-2030) Params: calendar (julian, coptic, ethiopic, buddhist, roc, islamic-umalqura, hebrew, persian, chin), date (Y-M-D in that calendar (Hebrew months numbered from Nisan))
- GET /calendars/weekday-names: Weekday names in a locale Params: locale (CLDR locale), context (format or stand-alone), width (wide, abbreviated, short or narrow), limit (Rows per page (1-200, default 50)), offset (Rows to skip for paging (default 0))
- GET /calendars/countries/{iso2}: A country's first day of week, weekend and the calendars it uses Params: iso2* (ISO 3166-1 alpha-2)

## Color Names: 30,000+ Named Colors

- id / MCP tool: colors
- base: https://api.lilambd.com/colors
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/colors/random?sample=1
- data as of: 2026-11-01; refresh: static snapshot; licence: MIT
- coverage: 30,242 names
- sources: color-name-list (meodai) (https://github.com/meodai/color-name-list)

Give every color a human name. Great for design tools, games, generative art, e-commerce filters and AI assistants.

- **30,242 named colors** with hex and RGB
- **Nearest name** for any hex (`/nearest?hex=ff8800`)
- Search by word (`ocean`, `sakura`, `rust`), exact lookup by name or hex
- Random palettes for games and prototypes

Source: color-name-list by meodai and contributors (MIT).

Routes:
- GET /colors/random: Random palette Params: count (How many (max 50))
- GET /colors/search: Search names Params: q (Word), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /colors/nearest: Nearest color name Params: hex* (Hex color without #), limit (How many (max 50))
- GET /colors/{name_or_hex}: Get a color Params: name_or_hex*

## Armed Conflict Events by Country and Region

- id / MCP tool: conflict
- base: https://api.lilambd.com/conflict
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/conflict/top?sample=1
- data as of: 2025-12-31; refresh: yearly (UCDP releases once a year); licence: CC BY 4.0
- coverage: events 1989-2025, aggregated by country and admin1 per year
- sources: UCDP Georeferenced Event Dataset (https://ucdp.uu.se/downloads/)
- known limits: Region totals include events only located to country precision, which inflates some regions.

Clean yearly conflict numbers for risk dashboards, travel-safety apps, news tools, research and AI agents.

- Country x year and region (state/province) x year, **1989-2025**
- Events, best-estimate deaths, civilian deaths
- Top countries or regions for any year, by deaths, civilian deaths or events
- ISO 3166-1 / ISO 3166-2 codes

Source: UCDP Georeferenced Event Dataset (CC BY 4.0). Attribution included in every response.

Note: UCDP events located only to the country level are placed at the country's central point, so the region containing that point can show inflated totals (e.g. Cherkasy in Ukraine). Use the country level for national totals.

Routes:
- GET /conflict/top: Top list Params: year (Year (default latest)), level (country | region), by (deaths | civilians | events), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /conflict/regions: Regions Params: country (ISO 3166-1 alpha-2), year (Year), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /conflict/countries/{iso2}: Country history Params: iso2*, from (First year), to (Last year), limit (Max rows (1-200, default 50)), offset (Rows to skip)

## How many deadly earthquakes, tsunamis and eruptions has this place seen since 1900?

- id / MCP tool: disaster-exposure
- base: https://api.lilambd.com/disaster-exposure
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/disaster-exposure/answer?lat=35.68&lon=139.77&sample=1
- data as of: 2026-11-01; refresh: with Lilambd/world-hazard rebuilds (monthly); licence: NOAA data: US Government work; compilation CC BY 4.0 Lilambd
- coverage: 6,288 records worldwide, 1900 to 2026
- sources: NOAA NCEI/WDS Global Significant Earthquake Database (https://www.ngdc.noaa.gov/hazel/view/hazards/earthquake/search); NOAA NCEI/WDS Global Historical Tsunami Database (https://www.ngdc.noaa.gov/hazel/view/hazards/tsunami/event-search); NOAA NCEI/WDS Significant Volcanic Eruptions Database (https://www.ngdc.noaa.gov/hazel/view/hazards/volcano/event-search)
- known limits: Only NOAA significant events (deaths, damage or large size). deaths_on_records sums published values and can double count a quake and its tsunami. Older records may have year or month only (day null). No floods, storms or droughts yet.

Give a point (lat, lon) and a radius. Returns NOAA NCEI significant earthquake, tsunami and volcanic eruption records within that radius since a chosen year: counts by kind, published deaths on the records, strongest quake magnitude and the five deadliest records with distance. Plus a per-decade table. Counts and distances computed by Lilambd from NOAA records; no personal data.

Routes:
- GET /disaster-exposure/: List routes
- GET /disaster-exposure/answer: Significant quakes, tsunamis and eruptions near a point since a year Params: lat* (Latitude, -90 to 90), lon* (Longitude, -180 to 180), radius_km (Search radius in km, 1 to 500 (default 100)), since (First year counted, 1900 or later (default 1900))
- GET /disaster-exposure/decades: Significant hazard records and deaths on records per decade since 1900 Params: kind (sig_earthquake, tsunami or volcano (default all)), limit (Rows, 1-200 (default 50)), offset (Rows to skip)

## Stock Exchanges and Trading Venues (MIC)

- id / MCP tool: exchanges
- base: https://api.lilambd.com/exchanges
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/exchanges/search?sample=1
- data as of: not recorded; refresh: static snapshot (ISO publishes monthly); licence: ISO 10383 public list
- coverage: 2,883 MIC entries
- sources: ISO 10383 Market Identifier Codes (https://www.iso20022.org/market-identifier-codes)
- known limits: Snapshot date of the MIC list not recorded; ISO updates the list monthly.

Look up any exchange or trading venue for fintech apps, trading bots and data pipelines.

- **2,883 venues** keyed by ISO 10383 Market Identifier Code (MIC) with operating MIC
- Name, acronym, city, country, market category (e.g. RMKT, MLTF, ATSS), operating/segment level
- Active or expired status with creation and close dates
- Filter by country, status, category; search by name, acronym or city

Source: ISO 10383 MIC list. Attribution included in every response.

Routes:
- GET /exchanges/{mic}: Get a venue Params: mic*
- GET /exchanges/search: Search venues Params: q (Name, acronym or city), country (ISO 3166-1 alpha-2), status (active | expired), category (Market category code), limit (Max rows (1-200, default 50)), offset (Rows to skip)

## Exposure brief: one finished security answer per company domain or tech stack

- id / MCP tool: exposure
- base: https://api.lilambd.com/exposure
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/exposure/brief?domain=fortinet.com&sample=1
- data as of: 2026-11-01; refresh: computed live on every call from ledgers refreshed every 2-24h; licence: per source; attribution in every response
- coverage: any domain or vendor/product stack; breach history for 989 organisations; ~7,000 regulator-reported incidents by organisation name; all 1,766 KEV CVEs; NVD 417,673 CVEs
- sources: CISA KEV (https://www.cisa.gov/known-exploited-vulnerabilities-catalog); FIRST EPSS (https://www.first.org/epss); NVD (https://nvd.nist.gov); Have I Been Pwned (https://haveibeenpwned.com); Lilambd JP disclosure ledger (https://lilambd.com); crt.sh Certificate Transparency (https://crt.sh)
- known limits: A brief is facts plus a ranking, not an audit. Stack matching by vendor/product names. Lookalike scan depends on crt.sh answering within 7 s (state shown). Breach history only where HIBP or the JP ledger has a record.

One call returns a ready answer about a company or a tech stack: its breach history (HIBP + Japan disclosures), every actively exploited CVE (CISA KEV) in the vendors/products you name (or the company's own products, inferred from the domain), which are used by ransomware and past CISA due dates, recent CVEs not yet exploited but with high EPSS, lookalike hostnames with new TLS certificates (Certificate Transparency, last 30 days), and a prioritised action list. Every item has a source URL; data_as_of per source. Facts only, no personal records.

Routes:
- GET /exposure/brief: Exposure brief for a domain and/or stack Params: domain*, stack (comma list of vendor or vendor:product, e.g. ivanti:connect_secure,microsoft:exc), lookalikes (0 to skip the certificate lookalike scan), limit (items per list (default 25))

## Who owns this company, where does it operate, what does it sell and research?

- id / MCP tool: firm-brief
- base: https://api.lilambd.com/firm-brief
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/firm-brief/answer?sample=1
- data as of: 2026-11-03; refresh: Snapshot of 2026-11-03; refreshed when the underlying ledgers are reloaded.; licence: GLEIF CC0; OpenAlex CC0; Open Food Facts ODbL 1.0 (attribution and share-alike apply to derived databases); tender notices are public EU / World Bank publications; Lilambd computed fields may be reused with attribution.
- coverage: 200 largest GLEIF ultimate-parent groups (by active descendants); 133 Open Food Facts brand owners with >=300 products; 196k OpenAlex institutions (36.6k companies); 2,237 tender notices (TED 2026-11-02, World Bank 2026-10-23..11-01); country stats for 241 countries; 543 country-to-parent-country ownership pairs.
- sources: GLEIF Golden Copy incl. relationship records (CC0) (https://www.gleif.org/en/lei-data/gleif-golden-copy); Open Food Facts / Beauty / Products data dumps (ODbL 1.0) (https://world.openfoodfacts.org/data); OpenAlex institutions and funders (CC0) (https://openalex.org); EU Tenders Electronic Daily notices (https://ted.europa.eu); World Bank procurement notices (https://projects.worldbank.org/en/projects-operations/procurement); EDINET code list (FSA Japan) and Wikidata (CC0), when loaded (https://disclosure2.edinet-fsa.go.jp)
- known limits: Name matching only (no fuzzy entity resolution): a brand-owner subsidiary (e.g. Nestle USA Inc.) is linked to its GLEIF group only by shared name. Groups outside the top 200, and Japanese groups searched in English (e.g. Toyota), have no GLEIF ownership block. Tenders are notices naming the buyer; award winners are not recorded, so tenders_won is always null. Flows are country-level counts of ownership links, not money. Brand search sees only each owner's top brand. OpenAlex name matches can include unrelated bodies sharing a word. EDINET and Wikidata blocks stay empty until those tables are loaded. No officers or other personal data.

One company brief joined from five open ledgers: GLEIF group ownership (ultimate parent, descendants, countries), Open Food Facts brand owners (brands, product categories, countries sold, with category market size), OpenAlex research organisations (sites, works, citations by country), public tender notices where the organisation is the buyer, and the home country's cross-border ownership links. Every block names its source and date; matching, country union and summaries are computed by Lilambd. Organisations only, no data about individual people.

Routes:
- GET /firm-brief/answer: Full company brief: ownership group, countries with evidence, brands and categories, research record, tenders, ownership flows of the home country Params: name (Company name (word-prefix, case- and accent-insensitive; Japanese/Chinese names ), lei (20-character Legal Entity Identifier (exact)), wikidata (Wikidata Q-id (exact), matched against OpenAlex institution records)
- GET /firm-brief/brands: Who owns this brand? Brand owner, its LEI, products, categories and countries sold Params: q* (Brand or owner name (punctuation and spaces ignored)), limit (Rows to return, default 50, max 200), offset (Rows to skip)
- GET /firm-brief/search: List firms covered (company groups, brand owners, research companies), filtered by name Params: q (Part of a name (word-prefix). Omit to list all groups and brand owners.), limit (Rows to return, default 50, max 200), offset (Rows to skip)
- GET /firm-brief/coverage: What the ledgers cover: counts per source and data dates

## World Natural Hazards and Warming

- id / MCP tool: hazard
- base: https://api.lilambd.com/hazard
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/hazard/gaps?sample=1
- data as of: 2026-11-01; refresh: static snapshot; reload when the Hugging Face shelf is updated; licence: Derived tables CC BY 4.0 (Lilambd); sources public domain (USGS, NOAA NCEI, Natural Earth, US Census) and CC BY 4.0 (OECD, World Bank, GeoNames)
- coverage: 4,596 admin-1 regions in 241 countries; 852 regions with M6+ since 1900; 1,240 regions with station warming; 5,924 GHCN-M v4 stations; 10,738 significant earthquakes, tsunamis and eruptions (NOAA NCEI); 41 joint-exposure regions; yearly anomaly 1951-2025
- sources: Lilambd World Hazard Ledger (Hugging Face) (https://huggingface.co/datasets/Lilambd/world-hazard); USGS ComCat earthquake catalogue (https://earthquake.usgs.gov/fdsnws/event/1/); NOAA NCEI Global Historical Hazard Database (significant earthquakes, tsunamis, volcanoes) (https://www.ngdc.noaa.gov/hazard/); NOAA NCEI GHCN-Monthly v4 (https://www.ncei.noaa.gov/products/land-based-station/global-historical-climatology-network-monthly); Natural Earth 10m admin-1 (https://www.naturalearthdata.com/); OECD regional statistics (https://www.oecd.org/regional/regional-statistics/); World Bank subnational population (https://datacatalog.worldbank.org/); GeoNames (https://www.geonames.org/)
- known limits: No region boundaries: place-risk picks the region of the nearest station within 30 km, else the nearest region centroid. No row for quakes = no M6+ within 1.5 deg since 1900; no warming = no station record, not no warming. 60 ISO 3166-2 codes are shared by several Natural Earth regions (e.g. PH-MNL); lookup returns the largest, use a1 ids for the rest. Some Natural Earth local names are wrong. Populations come from several sources and years; 332 regions have none. Raw M4.5+ quake list (313k rows, .csv.gz) not loaded.

How much each place shakes and how much it has warmed, in one API. Every number is counted or averaged from public measurement records; nothing is modelled.

- **Place risk** (finished answer): give lat/lon, get the admin-1 region, its quake level (M6+ and M7+ since 1900, density percentile), distance-weighted warming of nearby weather stations, significant quakes, tsunamis and eruptions within a radius with deaths, and a one-paragraph answer
- **Region profile** (finished answer): any of 4,596 states, provinces and prefectures by ISO 3166-2 (JP-13), a1 id, HASC or Wikidata id: quake counts and percentile, warming and percentile, station list, deadliest and latest events, joint-exposure flag
- **Regions**: search, filter by country, rank by M6+ density, M7+, warming, deaths or events, or find the nearest to a coordinate
- **Countries**: 241 countries with M6+/M7+ counts, M6+ per 1,000 km2, station warming, quake deaths, share of population in seismic regions, 2015-2024 temperature anomaly
- **Stations**: 5,924 GHCN-M v4 stations with 1961-1990 and 1991-2020 normals and warming; nearby search; 99.5% got warmer, mean +0.82 C
- **Events**: 10,738 NOAA NCEI significant earthquakes (6,634), tsunamis (2,673) and volcanic eruptions (903) with deaths, damage and houses destroyed; filter by kind, country, region, year, deaths, magnitude or radius
- **Joint exposure**: the 41 regions in the top quarter for both shaking and warming (218.7 million people)
- **Gaps**: populated regions with no station record, thin places first
- **Trends**: yearly temperature anomaly (arctic vs non-arctic, 1951-2025), M6/M7/M8 per decade, warming by latitude band
- **Features**: long-form indicators for countries and EU NUTS / US regions

Data as of 2026-11-01. Keys shared with other Lilambd APIs: iso2, ISO 3166-2, lat/lon, cell_0p5 = floor((lat+90)/0.5)*720 + floor((lon+180)/0.5). Sources: USGS ComCat (public domain), NOAA NCEI GHCN-M v4 and Global Historical Hazard Database (public domain), Natural Earth (public domain), populations from OECD and World Bank (CC BY 4.0), US Census (PD) and GeoNames (CC BY 4.0). Compiled by Lilambd (CC BY 4.0).

Routes:
- GET /hazard/gaps: Gaps Params: level (region | country), country (ISO 3166-1 alpha-2), min_population (Minimum population), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /hazard/events: Search disasters Params: kind (sig_earthquake | tsunami | volcano), country (ISO 3166-1 alpha-2), region (ISO 3166-2 or a1 id), from (First year), to (Last year), min_deaths (Minimum deaths), min_magnitude (Minimum magnitude), lat (Latitude), lon (Longitude), radius_km (Radius (default 300)), sort (newest | oldest | deaths | magnitude (nearest when lat/lon)), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /hazard/trends: Trends Params: band (all | arctic | non_arctic (empty = all bands))
- GET /hazard/regions: Search or rank regions Params: q (Name or ISO 3166-2), country (ISO 3166-1 alpha-2), sort (m6_density | m7 | m6 | warming | deaths | events | name), joint_exposure (true = only the 41 joint-exposure regions), lat (Latitude), lon (Longitude), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /hazard/summary: Totals
- GET /hazard/stations: Weather stations Params: country (ISO 3166-1 alpha-2), q (Station name), sort (warmest | coolest), lat (Latitude), lon (Longitude), radius_km (Radius (default 200)), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /hazard/countries: Rank countries Params: sort (m7 | m6 | m6_density | warming | deaths | deaths_per_million | pop_share_seismic), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /hazard/place-risk: Risk for a point (finished answer) Params: lat* (Latitude), lon* (Longitude), radius_km (Radius for stations and events (10-1000, default 300))
- GET /hazard/events/{id}: Get an event Params: id*, kind (sig_earthquake | tsunami | volcano)
- GET /hazard/regions/{id}: Region profile (finished answer) Params: id* (ISO 3166-2 code, a1 id, HASC or Wikidata id), events (Events per list (max 50))
- GET /hazard/features/{id}: Indicators Params: id*
- GET /hazard/stations/{id}: Get a station Params: id*
- GET /hazard/joint-exposure: Joint exposure Params: limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /hazard/countries/{iso2}: Country profile Params: iso2* (ISO 3166-1 alpha-2 or alpha-3)
- GET /hazard/regions/{id}/events: Events in a region Params: id*, kind (sig_earthquake | tsunami | volcano), sort (newest | deaths), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /hazard/regions/{id}/stations: Stations in a region Params: id*, limit (Max rows (1-200, default 50)), offset (Rows to skip)

## Japan Municipality Statistics

- id / MCP tool: japan
- base: https://api.lilambd.com/japan
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/japan/rank?indicator=A1101&sample=1
- data as of: not recorded; refresh: static snapshot; licence: CC BY 4.0 compatible (Government of Japan Standard Terms of Use 2.0)
- coverage: 1,741 municipalities, 122 indicators; years 2008 onward, some indicators are official projections up to 2070 (check the year field)
- sources: SSDSE (Statistics Center) (https://www.nstac.go.jp/use/literacy/ssdse/); Government of Japan open data (MIC, MHLW, MAFF, MLIT) (https://www.data.go.jp/)
- known limits: Each indicator has its own year; compare within one year.

Official statistics for every municipality in Japan in one JSON API, for real-estate tools, relocation apps, local-business dashboards and AI agents.

- **1,741 municipalities** by local government code (e.g. 13101 Chiyoda, 46218 Kirishima)
- **120+ indicators**: population and ageing, births/deaths, migration, households, municipal finance, businesses and employees, schools and pupils, hospitals, doctors, nursing care, vacant-house rate, residential land price, flood and landslide exposure, agriculture output, renewable energy (FIT), hometown tax (furusato nozei) 2008-2025
- Rank municipalities by any indicator, nationally or within a prefecture
- Indicator names are in Japanese (official wording); codes are stable

Sources: Japanese government open data under the Government Standard Terms of Use 2.0 (CC BY 4.0 compatible); non-commercial layers are excluded. Attribution included in every response.

Routes:
- GET /japan/rank: Rank municipalities Params: indicator* (Indicator code), year (Year (default latest)), prefecture (Prefecture code), order (desc | asc), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /japan/indicators: List indicators
- GET /japan/municipalities: List municipalities Params: q (Name (Japanese)), prefecture (Prefecture code), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /japan/municipalities/{code}: Municipality profile Params: code*, indicator (Only this indicator)

## Japan Hospitals and Prefecture Health Facts

- id / MCP tool: jp-health
- base: https://api.lilambd.com/jp-health
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/jp-health/mdc?sample=1
- data as of: 2025-03-31; refresh: static; reload when MHLW publishes the next DPC survey year; licence: Government of Japan Standard Terms of Use 2.0 (CC BY 4.0 compatible) for MHLW and e-Stat data; compilation by Lilambd CC BY 4.0
- coverage: Japan: 4,414 DPC-survey hospitals in 47 prefectures and 1,275 municipality codes; MDC breakdown for the 100 largest hospitals; 564 prefecture facts in 12 topics
- sources: MHLW DPC impact evaluation survey (discharge-patient survey) FY2024 (https://www.mhlw.go.jp/); Statistics Bureau of Japan, Statistics Dashboard (e-Stat) (https://dashboard.e-stat.go.jp/); Lilambd/japan-dpc-hospitals (Hugging Face, CC BY 4.0) (https://huggingface.co/datasets/Lilambd/japan-dpc-hospitals); Lilambd/japan-prefecture-facts (Hugging Face, CC BY 4.0) (https://huggingface.co/datasets/Lilambd/japan-prefecture-facts)
- known limits: MDC case breakdown only for the 100 largest hospitals; no hospital coordinates (no nearby search); names in Japanese only (no English or romanised names in the source); 281 hospitals without bed count; no population so no per-capita figures; facts mostly 2024, job ratio 2026-07; values as published.

Japan's hospital landscape in one API, built from the Ministry of Health, Labour and Welfare FY2024 DPC discharge-patient survey and official prefecture statistics.

- **4,414 hospitals** (every hospital in the FY2024 DPC survey: 1,765 DPC-participating, 225 preparing, 2,424 fee-for-service) with beds, annual discharges (11.38 million in total), DPC group and join year, JIS municipality code, prefecture (ja / en / ISO 3166-2) and national and prefectural rank
- **Search** by prefecture, municipality code, name, bed range, DPC status; sort by volume or beds
- **18 diagnostic categories (MDC)**: discharges and surgical cases per category for the 100 largest hospitals (1,800 rows), with top hospitals per category and per prefecture
- **Hospital profile** (finished answer): rank and percentile, share of prefecture and municipality discharges, category mix, other hospitals in the same municipality, similar-size peers, prefecture context
- **Prefecture health profile** (finished answer): hospitals, beds and discharges with national shares and ranks of 47, largest hospitals, top municipalities, bed-size bands, DPC status mix, category volumes, and minimum wage / real minimum wage / job ratio / regional price indexes / foreign residents with rank of 47
- **Rank and gaps**: prefectures ranked by 15 metrics; thin places first (fewest hospitals or beds, or most concentrated on one hospital)
- **564 prefecture facts** in 12 topics (2024; job ratio 2026-07)

Hospitals are organisations; no doctor or patient data. Names are as published in Japanese (name_ja). Values as published; ranks, shares and real minimum wage are derived by Lilambd. Sources: MHLW DPC survey FY2024 and Statistics Bureau of Japan e-Stat Statistics Dashboard, both Government of Japan Standard Terms of Use 2.0; compilation by Lilambd (CC BY 4.0).

Routes:
- GET /jp-health/mdc: Diagnostic categories
- GET /jp-health/gaps: Gaps Params: by (hospitals | beds | discharges | municipalities | large_hospitals | concentration), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /jp-health/rank: Rank prefectures Params: by (hospitals | beds | discharges | municipalities | hospitals_500_beds_plus | top10), order (desc | asc), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /jp-health/facts: Prefecture facts Params: topic (See /facts/topics), pref (Prefecture: code 01-47, ISO 3166-2 (JP-13), English name (Tokyo) or Japanese nam), sort (value | value_asc | (default pref)), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /jp-health/summary: Dataset summary
- GET /jp-health/hospitals: Search hospitals Params: pref (Prefecture: code 01-47, ISO 3166-2 (JP-13), English name (Tokyo) or Japanese nam), q (Text in the Japanese name), city_code (5-digit JIS municipality code), dpc_status (dpc_participating | dpc_preparing | fee_for_service), min_beds (Minimum beds), max_beds (Maximum beds), mdc (Diagnostic category M01-M18), min_cases (Minimum discharges in that category), sort (discharges | beds | name | smallest | id (with mdc: discharges | surgical | beds), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /jp-health/mdc/{mdc}: Top hospitals per category Params: mdc* (M01-M18), pref (Prefecture: code 01-47, ISO 3166-2 (JP-13), English name (Tokyo) or Japanese nam), sort (discharges | surgical | share), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /jp-health/prefectures: Prefectures Params: sort (hospitals | beds | discharges | concentration | real_min_wage | price_index | ..), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /jp-health/facts/topics: Fact topics
- GET /jp-health/cities/{city_code}: Hospitals in a municipality Params: city_code* (5-digit JIS code)
- GET /jp-health/prefectures/{pref}: Get a prefecture Params: pref* (01-47, JP-13, Tokyo or 東京都)
- GET /jp-health/hospitals/{hospital_id}: Get a hospital Params: hospital_id* (Hospital id)
- GET /jp-health/prefectures/{pref}/facts: Prefecture facts Params: pref* (Prefecture)
- GET /jp-health/hospitals/{hospital_id}/mdc: Hospital case mix Params: hospital_id* (Hospital id)
- GET /jp-health/prefectures/{pref}/hospitals: Hospitals in a prefecture Params: pref* (Prefecture), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /jp-health/hospitals/{hospital_id}/profile: Hospital profile Params: hospital_id* (Hospital id)
- GET /jp-health/prefectures/{pref}/health-profile: Prefecture health profile Params: pref* (01-47, JP-13, Tokyo or 東京都)

## Kinds: linked ledgers of kinds of things (instruments, woods, alloys, elements)

- id / MCP tool: kinds
- base: https://api.lilambd.com/kinds
- price: $0.05 USDC per call
- free sample: https://api.lilambd.com/kinds/chain?id=Q6607&sample=1
- data as of: 2026-11-01; refresh: monthly; licence: CC0-1.0
- coverage: musical instruments 2,392; woods 504; alloys 285; elements 174; instrument materials 41; links made_from 698, subclass_of 3,262
- sources: Wikidata (https://www.wikidata.org)
- known limits: Only what Wikidata records: many instruments lack materials or a Japanese name (591 of 2,392 have one); made_from is sparse; popularity = Wikipedia sitelinks count, which favours well-known Western instruments.

Kinds of things in one shared shape, linked to what they are made from: 2,392 musical instruments (Hornbostel-Sachs class, origin country), 504 woods, 285 alloys, 174 chemical elements and the materials instruments use. Follow made_from chains (guitar -> wood -> ...; cymbal -> bronze -> copper and tin). Japanese names where they exist. Source Wikidata (CC0).

Routes:
- GET /kinds/{id}: One kind with its links: made_from, subclass_of, used_in Params: id*
- GET /kinds/chain: What a kind is made from, step by step (up to 4 steps) Params: id*
- GET /kinds/facets: Which kinds ledgers exist and how big
- GET /kinds/search: Search kinds by facet and/or name (English or Japanese) Params: facet (musical_instrument, wood, alloy, element, instrument_material), q, limit, offset

## Koyomi: Calendars, Holidays and Helplines for AI Answers

- id / MCP tool: koyomi
- base: https://api.lilambd.com/koyomi
- price: $0.05 USDC per call
- free sample: https://api.lilambd.com/koyomi/day?sample=1
- data as of: not recorded; refresh: computed on request; holidays and helplines when updated; licence: Computed values CC0; holidays CC BY-SA 3.0 (date-holidays); helplines facts with source_url
- coverage: Calendars computed for 1900-2100 any place; holidays for countries in date-holidays; helplines for listed countries
- sources: astronomy-engine (computed); date-holidays; helplines: official pages, source_url per row
- known limits: Observational Hijri months may differ by a day locally; helplines checked on verified_on date; not advice. Crisis: /koyomi/helplines first, always free.

Data that grounds what people ask AI about feelings, fortune, love and daily life: computed calendars for any date and place (sexagenary cycle, rokuyo, Japanese and Chinese lunisolar dates, 24 solar terms, moon, sun sign, 10 world calendars), world public holidays, verified crisis and listening helplines, and a consult map.
Computed values CC0 (astronomy-engine MIT); holidays from date-holidays CC BY-SA 3.0; helplines carry a source_url per row. /helplines, /topics and /rules are always free.

Routes:
- GET /koyomi/day: Everything about one date Params: date (YYYY-MM-DD, default today), tz (UTC offset hours)
- GET /koyomi/sun: Sunrise, sunset, twilight, day length, moonrise/set for a place Params: lat*, lon*, date, tz
- GET /koyomi/range: Compact daily rows for up to 62 days, filter with has= Params: from*, to*, tz, has (e.g. 一粒万倍日, 大安, 立冬)
- GET /koyomi/rules [free]: FREE. How each value is computed
- GET /koyomi/topics [free]: FREE. What people ask AI, safety level, and which route grounds the answer
- GET /koyomi/holidays: Holidays of a country (and subdivision) for a year Params: country*, year, subdiv, lang, types
- GET /koyomi/countries: Countries with holiday rules; ?country=XX lists subdivisions Params: country
- GET /koyomi/helplines [free]: FREE. Crisis and listening lines with hours, languages, source_url, verified_on Params: country
- GET /koyomi/moon-phases: Principal moon phases of a year Params: year*
- GET /koyomi/solar-terms: 24 solar terms of a year with exact UTC time Params: year*

## Land Use and Forest Loss by Region

- id / MCP tool: land
- base: https://api.lilambd.com/land
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/land/gaps?sample=1
- data as of: 2026-11-01; refresh: yearly (new Hansen GFC release); licence: CC BY 4.0 (derived tables) with source attribution; NOAA and Natural Earth public domain; ENVO CC0
- coverage: 4,596 admin-1 regions and 251 countries; forest loss 2001-2025 (canopy 10/30/50%), land cover 2000-2020, built-up 1975-2025, surface water 1984-2021, soil (static); ground exposure for 4,182 regions
- sources: Lilambd World Earth Surface Ledger (Hugging Face) (https://huggingface.co/datasets/Lilambd/world-earth); Lilambd World Ground Ledger (Hugging Face) (https://huggingface.co/datasets/Lilambd/world-ground); Hansen/UMD/Google/USGS/NASA Global Forest Change GFC-2025-v1.13 (https://storage.googleapis.com/earthenginepartners-hansen/GFC-2025-v1.13/download.html); GLAD Global Land Cover and Land Use Change 2000-2020 v2 (https://storage.googleapis.com/earthenginepartners-hansen/GLCLU2000-2020/v2/download.html); JRC Global Surface Water v1.4 (https://global-surface-water.appspot.com/); JRC GHSL GHS-BUILT-S and GHS-POP R2023A (https://ghsl.jrc.ec.europa.eu/); ISRIC SoilGrids 2.0 (https://soilgrids.org/); NOAA NCEI Global Volcano database (https://www.ngdc.noaa.gov/hazard/volcano.shtml); Macrostrat (https://macrostrat.org/); Environment Ontology (ENVO) (https://www.ebi.ac.uk/ols4/ontologies/envo); Natural Earth (https://www.naturalearthdata.com/)
- known limits: Forest loss is stand-replacing canopy loss (fire, logging, storms), not net deforestation; gain only 2000-2012. Loss-by-year kept for canopy 10, 30 and 50% only. 2025 built-up is a GHSL nowcast. Soil is static SoilGrids model output. Region centres (nearby, cell_0p5) only for the 4,182 regions in World Ground Ledger. 60 ISO 3166-2 codes are shared by more than one Natural Earth region (largest wins; use a1). Plate-boundary exposure fields excluded (PB2002 licence not stated). Some Natural Earth local names are wrong.

What the ground of every state, province and prefecture is made of, and how it changed, counted from 30 m satellite pixels. One API for site screening, ESG and deforestation checks, urban growth, agriculture and climate apps and AI agents.

- **Region profile** (finished answer): one call returns a region's forest loss per year 2001-2025, land cover 2000-2020 with what turned into what, built-up area and population 1975-2025, surface water gained and lost 1984-2021, soil properties, people living near the coast and volcanoes, its rank inside its country and the world, and a plain-English headline
- **Regions**: 4,596 admin-1 regions with ISO 3166-2, Wikidata, HASC, English, local and Japanese names, area and population; search, filter by country, **nearby** by lat/lon
- **Forest**: tree cover in 2000 and loss for every year 2001-2025 at canopy thresholds 10, 30 and 50% (base totals at 7 thresholds) (273,293 region-year rows)
- **Land cover**: 11 cover groups in 2000, 2005, 2010, 2015 and 2020 and the 2000->2020 transition matrix
- **Built-up**: GHSL built-up km2 and population every 5 years 1975-2025 (2025 is the GHSL nowcast, flagged)
- **Surface water**: JRC transition classes (permanent, new, lost, seasonal...) per region
- **Soil and ground**: SoilGrids properties at two depths, people within 10 km of the coast and 50 km of an erupting volcano
- **Rankings**: fastest forest loss, most recent loss, fastest built-up growth, most water lost, cropland gain and more, for regions or countries
- **Gaps**: thin places first, so missing data is visible
- **World** totals by year, cover-group legend mapped to ENVO terms, and 1,780 country and 9,516 NUTS / US region features

World forest loss (canopy >= 30%) in 2025: 255,621 km2. Nothing is generated by a model; every number is a pixel count or area-weighted mean. Data as of 2026-11-01.

Sources (all commercial use allowed with attribution): Hansen/UMD/Google/USGS/NASA Global Forest Change GFC-2025-v1.13 (CC BY 4.0); GLAD Global Land Cover and Land Use Change 2000-2020 v2 (CC BY 4.0); JRC Global Surface Water v1.4 (CC BY 4.0); JRC GHSL R2023A (CC BY 4.0); ISRIC SoilGrids 2.0 (CC BY 4.0); NOAA NCEI volcanoes (public domain); Macrostrat (CC BY 4.0); ENVO (CC0); Natural Earth (public domain). Compiled by Lilambd.

Routes:
- GET /land/gaps: Gaps Params: country (ISO 3166-1 alpha-2 or alpha-3), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /land/world: World totals by year Params: thr (Tree canopy threshold in % for the loss-by-year series: 10, 30 or 50 (default 30)
- GET /land/nearby: Regions near a point Params: lat* (Latitude), lon* (Longitude), radius_km (Radius in km (default 300, max 2000)), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /land/classes: Legend and sources Params: detail (true adds the 256 raw GLCLU classes)
- GET /land/regions: Search regions Params: country (ISO 3166-1 alpha-2 or alpha-3), q (Name, local name, Japanese name or ISO code), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /land/features: Country and NUTS / US region features Params: country (ISO 3166-1 alpha-2 or alpha-3), region_id (NUTS:xx or US:xx id), feature (Feature name), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /land/rankings: Rankings Params: by (forest_loss_share | forest_loss_km2 | recent_forest_loss | forest_loss_accelerat), level (region | country), country (ISO 3166-1 alpha-2 or alpha-3), min_area_km2 (Minimum area (default 500 for forest_loss_share)), min_tree_cover_km2 (Minimum tree cover 2000 for forest_loss_share (default 100)), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /land/countries: Countries Params: limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /land/regions/{id}: Get a region Params: id* (ISO 3166-2 code (e.g. JP-13, BR-PA, US-CA), an a1 id (e.g. BRA-594) or a Wikidat)
- GET /land/countries/{code}: Country profile Params: code* (ISO 3166-1 alpha-2 or alpha-3), thr (Tree canopy threshold in % for the loss-by-year series: 10, 30 or 50 (default 30)
- GET /land/regions/{id}/soil: Soil and ground exposure Params: id* (ISO 3166-2 code (e.g. JP-13, BR-PA, US-CA), an a1 id (e.g. BRA-594) or a Wikidat)
- GET /land/regions/{id}/built: Built-up and population 1975-2025 Params: id* (ISO 3166-2 code (e.g. JP-13, BR-PA, US-CA), an a1 id (e.g. BRA-594) or a Wikidat)
- GET /land/regions/{id}/water: Surface water change Params: id* (ISO 3166-2 code (e.g. JP-13, BR-PA, US-CA), an a1 id (e.g. BRA-594) or a Wikidat)
- GET /land/regions/{id}/forest: Forest loss by year Params: id* (ISO 3166-2 code (e.g. JP-13, BR-PA, US-CA), an a1 id (e.g. BRA-594) or a Wikidat), thr (Tree canopy threshold in % for the loss-by-year series: 10, 30 or 50 (default 30)
- GET /land/regions/{id}/profile: Region profile Params: id* (ISO 3166-2 code (e.g. JP-13, BR-PA, US-CA), an a1 id (e.g. BRA-594) or a Wikidat), thr (Tree canopy threshold in % for the loss-by-year series: 10, 30 or 50 (default 30)
- GET /land/regions/{id}/land-cover: Land cover 2000-2020 Params: id* (ISO 3166-2 code (e.g. JP-13, BR-PA, US-CA), an a1 id (e.g. BRA-594) or a Wikidat), min_km2 (Smallest transition to list in km2)

## World Languages (Glottolog)

- id / MCP tool: languages
- base: https://api.lilambd.com/languages
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/languages/search?sample=1
- data as of: not recorded; refresh: static snapshot; licence: CC BY 4.0
- coverage: 8,120 languages
- sources: Glottolog 5 (https://glottolog.org)
- known limits: Exact Glottolog minor version not recorded.

Language metadata for localisation tools, language-learning apps, games and AI pipelines.

- **8,120 languages** with Glottocode and ISO 639-3
- Language family, countries where spoken, autonym where known
- Endangerment status (not_endangered, threatened, shifting, moribund, nearly_extinct, extinct)
- Filter by country, family, endangerment; look up by `jpn` or `nucl1643`

Source: Glottolog 5 (CC BY 4.0). Attribution included in every response.

Routes:
- GET /languages/search: Search languages Params: q (Name or autonym), country (ISO 3166-1 alpha-2), family (Family glottocode or name), endangerment (Status), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /languages/{code}: Get a language Params: code*

## Open Source Licenses (SPDX)

- id / MCP tool: licenses
- base: https://api.lilambd.com/licenses
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/licenses/search?sample=1
- data as of: 2026-11-01; refresh: static snapshot; licence: CC0
- coverage: 740 licences
- sources: SPDX License List (https://spdx.org/licenses/)

License metadata for dependency scanners, SBOM tools, package registries and AI coding agents.

- **740 SPDX licenses** with official id (`MIT`, `Apache-2.0`, `GPL-3.0-or-later`)
- OSI-approved and FSF-libre flags, deprecated flag
- Reference and see-also links

Source: SPDX License List (CC0).

Routes:
- GET /licenses/{id}: Get a license Params: id*
- GET /licenses/search: Search licenses Params: q (Id or name), osi_approved (true | false), limit (Max rows (1-200, default 50)), offset (Rows to skip)

## Local Markets: markets away from big cities

- id / MCP tool: local-markets
- base: https://api.lilambd.com/local-markets
- price: $0.05 USDC per call
- free sample: https://api.lilambd.com/local-markets/near?lat=-1.29&lon=36.82&sample=1
- data as of: 2026-10-21; refresh: with each Overture release; licence: CDLA-Permissive-2.0 (places); locality values CC0
- coverage: 137,975 market places worldwide; 58,804 regional or remote
- sources: Overture Maps Foundation places (https://overturemaps.org); Natural Earth populated places (https://www.naturalearthdata.com)
- known limits: Overture under-records China and much of Africa; locality is distance to the nearest Natural Earth city (3,089 cities), not an official rural classification; farmers_market can include farm shops.

Every recorded market place in the world tagged by how far it sits from cities (Natural Earth): big_city (<=30 km of a 1M+ city), city (<=30 km of a 100k+ city), regional (30-100 km), remote (>100 km). Default answers return only regional and remote markets, the village and rural markets big catalogues bury. Places: Overture Maps Foundation (CDLA-Permissive-2.0). Cities: Natural Earth (public domain). Locality computed by Lilambd.

Routes:
- GET /local-markets/near: Local markets near a point, nearest first, with distance_km Params: lat*, lon*, radius_km, locality (comma list of big_city, city, regional, remote), limit, offset
- GET /local-markets/{id}: One market with locality and distances Params: id*
- GET /local-markets/search: Search local markets by name Params: q*, locality, limit
- GET /local-markets/country: Local markets of one country Params: country*, locality, limit, offset
- GET /local-markets/countries: Per country: local, remote and all recorded markets, local share

## World Markets: Farmers, Flea, Night and Fish Markets

- id / MCP tool: markets
- base: https://api.lilambd.com/markets
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/markets/kinds?sample=1
- data as of: 2026-10-21; refresh: monthly (Overture releases monthly; refresh planned); licence: CDLA-Permissive-2.0 / Apache-2.0; Natural Earth public domain
- coverage: 137,975 places in 207 countries (release 2026-10-21.0)
- sources: Overture Maps places (https://docs.overturemaps.org/); Natural Earth (https://www.naturalearthdata.com/)
- known limits: Coverage is thin in some countries (e.g. China); /gaps shows where.

Find real-world markets anywhere. For travel apps, local-discovery apps, maps, games and AI agents.

- **137,975 places** in 207 countries, 13 kinds: farmers_market, flea_market, auction_house, event_market_venue, night_market, trade_fair_venue, fish_market, public_market, street_stall, flower_market, seasonal_market, bazaar_souk, hawker_centre
- Name, country, region, town, coordinates, website, open/closed status
- **Nearby** markets for any coordinate with distance
- **Countries**: markets per country and per million people, with population and GDP per capita, to spot under-covered places
- Free sample: huggingface.co/datasets/Lilambd/world-markets

Source: Overture Maps Foundation places, release 2026-10-21.0 (CDLA-Permissive-2.0; parts Apache-2.0). Coverage reflects Overture's sources and is thin in some countries (e.g. China).

Routes:
- GET /markets/{id}: Get a market Params: id*
- GET /markets/kinds: Market kinds
- GET /markets/nearby: Nearby markets Params: lat* (Latitude), lon* (Longitude), kind (Market kind (see /kinds)), include_closed (true to include permanently closed), limit (How many (max 50))
- GET /markets/search: Search markets Params: q (Name or town), country (ISO 3166-1 alpha-2), kind (Market kind), include_closed (true | false), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /markets/countries: Markets by country Params: country (ISO 3166-1 alpha-2 (empty = all)), kind (Market kind), min_population (Only countries above this population), sort (population | per_million (lowest first)), limit (Max rows (1-200, default 50)), offset (Rows to skip)

## Which languages should I localize into next?

- id / MCP tool: next-languages
- base: https://api.lilambd.com/next-languages
- price: $0.05 USDC per call
- free sample: https://api.lilambd.com/next-languages/answer?sample=1
- data as of: 2026-11-03; refresh: manual; next rebuild with CLDR release; licence: Data: Unicode License v3 (CLDR, UCD), CC BY 4.0 (Glottolog). Scores computed by Lilambd.
- coverage: 60 candidate languages (CLDR languages with territory data, excluding en ja zh es pt fr de ko hi ar); 173 Unicode-encoded scripts
- sources: Unicode CLDR (cldr-json) (https://github.com/unicode-org/cldr-json); Unicode Character Database 17.0 Scripts.txt (https://www.unicode.org/Public/17.0.0/ucd/Scripts.txt); ISO 15924 (npm iso-15924) (https://www.unicode.org/iso15924/); Glottolog 5 CLDF (https://github.com/glottolog/glottolog-cldf); Lilambd world-marketplaces v3 (digital marketplaces per country) (https://huggingface.co/datasets/Lilambd/world-marketplaces)
- known limits: Speakers include second-language users. GDP is PPP from CLDR. The 10 already-covered languages are never ranked. Payment or sanctions constraints in target countries are not considered.

Ranks 60 candidate languages for product localization beyond en ja zh es pt fr de ko hi ar, by CLDR speakers (L1+L2), GDP (PPP) reach, digital-marketplace reach, official-language countries and CLDR locale coverage. Also lists writing systems (ISO 15924) with Unicode 17.0 character counts, direction and speakers.

Routes:
- GET /next-languages/: Discovery: routes and method
- GET /next-languages/answer: Next languages to localize into, ranked Params: have (Comma-separated CLDR or ISO 639-3 codes you already support; they are skipped.), limit (Rows to return, default 50, max 200.), offset (Rows to skip.)
- GET /next-languages/scripts: Writing systems by speakers Params: rtl (true = right-to-left scripts only, false = left-to-right only.), limit (Rows to return, default 50, max 200.)
- GET /next-languages/script/{code}: One writing system by ISO 15924 code Params: code* (ISO 15924 four-letter code.)

## World Organizations: Research Bodies Societies Funders and Publishers

- id / MCP tool: orgs
- base: https://api.lilambd.com/orgs
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/orgs/gaps?sample=1
- data as of: 2026-10-14; refresh: static snapshot; licence: CC0 (OpenAlex); US government work (IRS)
- coverage: 195,967 OpenAlex organisations (snapshot 2026-10-14) + 606,825 US nonprofits
- sources: OpenAlex (https://openalex.org); IRS Exempt Organizations Business Master File (https://www.irs.gov/charities-non-profits/exempt-organizations-business-master-file-extract-eo-bmf)
- known limits: Organisation form is inferred from the name. OpenAlex coverage follows scholarly records. /nonprofits covers US tax-exempt organisations only (IRS EO BMF eo1-4 as fetched 2026-11-01, 606,825 rows); care-of names and street addresses are dropped on purpose; NTEE cause is missing for about 196,000 rows; revenue and assets are the latest filed figures.

A world directory of organizations that are not just companies or governments. For research tools, grant finders, CRM enrichment, maps, education apps and AI agents.

- **195,967 organizations**: 143,455 research-linked institutions (education, nonprofit, healthcare, facility, government, archive, company...), 41,799 research funders, 10,713 publishers
- **Form** inferred from the name: society, association, academy, foundation, trust_or_charity, federation_or_network, council_or_agency, museum_library_archive, hospital_or_clinic, university_or_school, institute_or_center
- Coordinates, city, region, country, homepage, ROR and Wikidata ids, output counts (works, citations, h-index)
- **Nearby** organizations for any coordinate
- **Gaps**: organizations per million people by country, lowest first, so thinly recorded places are not left out
- **US nonprofits**: 606,825 tax-exempt organizations (445,659 public charities, 42,699 private foundations, 118,467 other exempt groups) with town, state, ZIP, NTEE cause, ruling year, assets and revenue. Sort by smallest revenue to find small local groups. Care-of names and street addresses are left out on purpose.

Sources: OpenAlex (CC0) snapshot 2026-10-14, with ROR ids (CC0); IRS Exempt Organizations Business Master File (public domain). OpenAlex coverage follows scholarly records; small local groups are better covered by the nonprofits endpoints.

Routes:
- GET /orgs/gaps: Thinly recorded countries Params: kind (Kind), form (Form), min_population (Only countries above this population), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /orgs/{id}: Get an organization Params: id*
- GET /orgs/forms: Forms
- GET /orgs/kinds: Kinds
- GET /orgs/nearby: Nearby organizations Params: lat* (Latitude), lon* (Longitude), kind (Kind), form (Form), limit (How many (max 50))
- GET /orgs/search: Search organizations Params: q (Name or acronym), country (ISO 3166-1 alpha-2), kind (education | nonprofit | healthcare | facility | government | archive | company |), form (society | association | academy | foundation | ...), source (institutions | funders | publishers), sort (works | name), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /orgs/countries: Organizations by country Params: country (ISO 3166-1 alpha-2 (empty = all)), kind (Kind), form (Form), sort (population | per_million), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /orgs/nonprofits/{ein}: Get a US nonprofit Params: ein*
- GET /orgs/nonprofits/causes: US nonprofits by cause Params: state (US state code)
- GET /orgs/nonprofits/search: Search US nonprofits Params: q (Name text), state (US state code), city (Town (exact, case-insensitive)), postcode (5-digit ZIP), kind (charity | private_foundation | other_exempt_org), cause (NTEE major letter, e.g. K = food and agriculture), min_revenue (Minimum revenue (USD)), max_revenue (Maximum revenue (USD)), sort (revenue | smallest | name | oldest), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /orgs/nonprofits/states: US nonprofits by state Params: cause (NTEE major letter), sort (most | fewest)

## Pet entry: can my dog or cat cross this border in time?

- id / MCP tool: pet-entry
- base: https://api.lilambd.com/pet-entry
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/pet-entry/answer?to=JP&sample=1
- data as of: 2026-11-01; refresh: weekly recheck of source pages; licence: facts from public government pages; Lilambd shaping CC BY 4.0
- coverage: 5 destinations (JP, GB, EU, US, SG), 11 routes; origin lists: Japan designated regions, Singapore Schedule I, CDC 110 high-risk countries
- sources: Government pages per destination (MAFF AQS Japan, GOV.UK, European Commission DG SANTE, US CDC, NParks AVS Singapore)
- known limits: v0: 5 destinations. Australia not yet (site blocks automated reading). EU Annex I-II list and Singapore Schedule II list not loaded, so those answers return both possible routes. GB blood-test country list not on the main page. lead_days_from_scratch is computed by Lilambd. Not veterinary or legal advice.

Official rules for taking a pet dog, cat or ferret into another country: microchip, rabies shots, blood test and waiting days, advance notice, papers, quarantine, and the shortest days from scratch to legal entry. Every row cites the government page it was read from. Facts only, not veterinary or legal advice.

Routes:
- GET /pet-entry/answer: One finished answer for a move: route, verdict for your days left, steps in order, quarantine, sources Params: to* (Destination ISO2 (any EU member maps to EU)), from (Where the pet lives now (ISO2)), species (dog, cat or ferret), days (Days until arrival), vaccinated_in (US only: us or foreign)
- GET /pet-entry/country: All routes for one destination Params: to*
- GET /pet-entry/countries [free]: Which destinations are covered

## What natural hazards and infrastructure surround this latitude/longitude?

- id / MCP tool: place-brief
- base: https://api.lilambd.com/place-brief
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/place-brief/answer?lat=35.68&lon=139.76&sample=1
- data as of: 2026-11-01; refresh: Hazard catalogues fetched 2026-11-01 (static snapshot, refresh planned); airports snapshot 2026-11-01; markets with each Overture release; power plants frozen (source ended 2021).; licence: NOAA/USGS US Government public domain; Natural Earth and OurAirports public domain; WRI GPPD CC BY 4.0; Overture places CDLA-Permissive-2.0; World Bank CC BY 4.0. Computed fields (distances, counts, ranks) by Lilambd.
- coverage: Any lat/lon on Earth. 10,738 NCEI hazard events (6,697 significant earthquakes, 3,138 tsunamis, 903 volcanic eruptions), 852 admin1 M6+ rates, 5,924 GHCN stations, 4,596 admin1 regions, 3,089 cities, 86,329 airports, 34,936 power plants, 137,975 market places. Cyclones (IBTrACS), recent-event feed, seaports and world-cities profiles are wired in and appear automatically once those loads finish.
- sources: NOAA NCEI/WDS Global Significant Earthquake Database (https://www.ngdc.noaa.gov/hazel/view/hazards/earthquake/search); NOAA NCEI/WDS Global Historical Tsunami Database (https://www.ngdc.noaa.gov/hazel/view/hazards/tsunami/event-search); NOAA NCEI/WDS Significant Volcanic Eruptions Database (https://www.ngdc.noaa.gov/hazel/view/hazards/volcano/event-search); USGS ComCat earthquake catalog (M6+ 1900-2026, aggregated per admin1) (https://earthquake.usgs.gov/fdsnws/event/1/); NOAA GHCN-Monthly v4 (station normals 1961-1990, 1991-2020) (https://www.ncei.noaa.gov/pub/data/ghcn/v4/); Natural Earth admin-1 and populated places (https://www.naturalearthdata.com/); OurAirports (https://ourairports.com/data/); WRI Global Power Plant Database v1.3.0 (https://datasets.wri.org/dataset/globalpowerplantdatabase); Overture Maps places release 2026-10-21.0 (https://docs.overturemaps.org/); World Bank World Development Indicators (country population, GDP per capita, income group) (https://data.worldbank.org/)
- known limits: Historical record only, not a hazard model or probability. NCEI 'significant' catalogues list damaging or large events, not all quakes. Tsunamis are counted by source location, not run-up on this coast. Admin1 is inferred from the nearest polygon-tagged station/event within 40 km, else nearest admin1 centroid (can be wrong near borders; null far offshore). No soil, geology, flood-zone or elevation model yet: elevation shown is the nearest airport's. Power plants miss anything built after 2020. Markets under-recorded in China and much of Africa. Cyclones, seaports, recent events and city profiles show 'not loaded yet' until their loads complete.

One finished place brief for any point on Earth: historical significant earthquakes, tsunami sources and volcanic eruptions near the point (NOAA NCEI), the admin1 region's USGS M6+ rate since 1900 and its rank, nearby GHCN station warming (1991-2020 vs 1961-1990), country hazard totals, nearest cities, scheduled airports, power plants (capacity by fuel), market places and ports, each with great-circle distance in km. Every block names its source and data date; computed fields are marked as computed by Lilambd. Historical record only, not a hazard model, engineering or legal advice.

Routes:
- GET /place-brief/answer: Finished place brief for one point: hazards record, regional rates, nearest cities, airports, power, markets, with km Params: lat* (Latitude, decimal degrees WGS84 (-90..90)), lon* (Longitude, decimal degrees WGS84 (-180..180)), radius_km (Search radius for earthquakes and tsunami sources, km (default 150, 10..500))
- GET /place-brief/nearest: Nearest features of one kind to a point, sorted by distance in km Params: lat* (Latitude, decimal degrees), lon* (Longitude, decimal degrees), kind* (What to list), radius_km (Search radius km (default 200, max 2000)), scheduled (kind=airport only: scheduled passenger service only), include_heliports (kind=airport only: also list heliports and other non-airfield types), limit (Rows to return (default 50, max 200)), offset (Rows to skip)
- GET /place-brief/coverage: Free discovery: which layers are covered, row counts, years and sources

## Online Platforms Directory

- id / MCP tool: platforms
- base: https://api.lilambd.com/platforms
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/platforms/search?sample=1
- data as of: not recorded; refresh: static snapshot; licence: CC0
- coverage: 5,001 platforms
- sources: Wikidata (https://www.wikidata.org)
- known limits: Wikidata snapshot date not recorded.

A categorised directory of the world's online platforms for market research, competitor maps, onboarding flows and AI agents.

- **5,001 platforms** with Wikidata id, website, country, founding/dissolution date, ultimate parent company
- 20+ categories: app_store, social, streaming, online_retail, marketplace, payment, food_delivery, ride_hailing, dating, jobs, crypto_exchange, lending_crowdfunding, travel, review, ticketing...
- Filter by category and country, search by name or domain

Source: Wikidata (CC0), categorised by Lilambd.

Routes:
- GET /platforms/search: Search platforms Params: q (Name or domain), category (Category), country (ISO 3166-1 alpha-2), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /platforms/categories: List categories
- GET /platforms/{wikidata}: Get a platform Params: wikidata*

## Power Plants of the World

- id / MCP tool: power
- base: https://api.lilambd.com/power
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/power/fuels?sample=1
- data as of: 2021-06-01; refresh: none (source no longer updated); licence: CC BY 4.0
- coverage: 34,936 plants; v1.3.0 released June 2021; commissioning years up to 2020
- sources: WRI Global Power Plant Database v1.3.0 (https://datasets.wri.org/dataset/globalpowerplantdatabase)
- known limits: Source frozen in 2021: plants built after 2020 are missing.

Global power plant lookup for energy apps, climate tools, maps, simulations, city-building games and AI agents.

- **34,936 plants**, 167 countries, 5.7 TW total capacity
- Capacity (MW), primary and other fuels (coal, gas, hydro, nuclear, solar, wind...), coordinates, owner, commissioning year
- Reported or estimated annual generation (GWh) for ~11,400 plants
- **Nearby** plants for any coordinate, **country fuel mix**, filter by fuel and size

Source: WRI Global Power Plant Database v1.3.0 (CC BY 4.0), data as of 2021.

Routes:
- GET /power/{id}: Get a plant Params: id*
- GET /power/fuels: Fuels
- GET /power/nearby: Nearby power plants Params: lat* (Latitude), lon* (Longitude), fuel (Primary fuel), min_mw (Minimum capacity MW), limit (How many (max 50))
- GET /power/search: Search plants Params: q (Name or owner), country (ISO 3166-1 alpha-3), fuel (Primary fuel), min_mw (Minimum capacity MW), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /power/countries: Country fuel mix Params: country (ISO 3166-1 alpha-3 (empty = all))

## World Prices by Country

- id / MCP tool: prices
- base: https://api.lilambd.com/prices
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/prices/gaps?sample=1
- data as of: 2026-11-02; refresh: manual, with the Hugging Face shelf (commodities monthly, inflation database when the World Bank updates); licence: CC BY 4.0 (World Bank, Eurostat); public domain (US EIA)
- coverage: 70 commodity prices and 16 indices monthly 1960-01 to 2026-08; headline, food and energy CPI for 186 countries monthly 1970 to 2025-03; Eurostat household electricity and gas yearly (about 40 European countries); EIA residential energy prices for 50 US states and DC yearly to 2024
- sources: World Bank Commodity Price Data (Pink Sheet) (https://www.worldbank.org/en/research/commodity-markets); World Bank Global Database of Inflation (https://www.worldbank.org/en/research/brief/inflation-database); Eurostat household energy prices (nrg_pc_202, nrg_pc_204) (https://ec.europa.eu/eurostat/web/energy/database); US EIA State Energy Data System (https://www.eia.gov/state/seds/); Lilambd World Prices Long Record (https://huggingface.co/datasets/Lilambd/world-prices)
- known limits: CPI values are index levels with country-specific base years, compare changes only; World Bank inflation database monthly only to 2025-03; commodity prices are world benchmarks, not local retail prices; some countries (e.g. Australia) have no monthly CPI; Kosovo appears in Eurostat data only

What things cost and how fast prices move, by country and by commodity, in one API.

- **World commodity prices**: 70 benchmark prices (crude oil, natural gas, coal, rice, wheat, maize, coffee, cocoa, sugar, beef, fertilisers, copper, gold...) and 16 World Bank price indices, monthly from 1960-01 to 2026-08, with latest value, change against a year and five years before, and the 12-month low and high
- **Consumer prices for 186 countries**: headline, food and energy CPI monthly from 1970 to 2025-03, with year-on-year change and yearly inflation
- **Rankings**: fastest-rising food, energy or headline prices; biggest food price rise 2021-2023; most unstable inflation; dearest and cheapest household electricity and gas in about 40 European countries (Eurostat) and residential energy prices in the 50 US states and DC (EIA)
- **Compare** up to 12 countries side by side for one item
- **Country profile**: a finished answer for one country: current inflation, five-year change, rank, how strongly and how late food and energy prices follow world commodity prices, European household energy prices, and the world drivers right now, with a plain-English summary
- **Gaps**: countries with the thinnest or oldest price data first

Measured: 220,053 monthly and 26,898 yearly observations, 94 items. CPI values are index levels with country-specific base years, so the API compares changes, not levels, across countries. World Bank inflation data are monthly up to 2025-03; nothing is estimated or filled.

Sources: World Bank Commodity Price Data (Pink Sheet, CC BY 4.0); World Bank Global Database of Inflation (Ha, Kose and Ohnsorge 2023, CC BY 4.0); Eurostat household energy prices (CC BY 4.0); US EIA State Energy Data System (public domain). Compiled by Lilambd.

Routes:
- GET /prices/gaps: Data gaps Params: limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /prices/rank: Rank countries Params: item (cpi_food | cpi_energy | cpi_headline | px_eu_hh_elec | px_eu_hh_gas | prc_food_c), by (yoy | change_5y | level | change), order (desc | asc), year (Year for yearly items), continent (Continent filter), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /prices/items: Items Params: q (Text in name or code), kind (commodity | index | cpi | household_energy_eu | us_state_energy), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /prices/compare: Compare countries Params: countries* (Comma-separated ISO2 codes), item (Item code), from (First year), to (Last year)
- GET /prices/summary: Dataset summary
- GET /prices/countries: Countries Params: q (Name, ISO2 or ISO3), continent (Continent), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /prices/us-states: US state energy prices Params: item (px_us_elec_residential | px_us_gas_residential | px_us_energy_avg), year (Year), order (desc (dearest) | asc (cheapest)), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /prices/commodities: Latest world commodity prices Params: kind (commodity | index | all), q (Text), sort (name | yoy (fastest rising) | falling | change_5y), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /prices/items/{item}: One item Params: item*
- GET /prices/countries/{iso2}: Country price series Params: iso2*, item (cpi_headline | cpi_food | cpi_energy | px_eu_hh_elec | px_eu_hh_gas), freq (monthly | yearly), from (First year), to (Last year)
- GET /prices/commodities/{item}: Commodity price history Params: item*, freq (monthly | yearly), from (First year), to (Last year)
- GET /prices/countries/{iso2}/profile: Country price profile Params: iso2*

## World Pulse Daily Tech and Money Briefing

- id / MCP tool: pulse
- base: https://api.lilambd.com/pulse
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/pulse/fx?sample=1
- data as of: 2026-11-03; refresh: daily from the HF shelf (pushed ~05:45 UTC) via api_pub.pl_load_day; licence: Census CC BY-SA 4.0 (Lilambd); sources CC0 / CC BY-SA / ECB reuse with source / public API facts
- coverage: 13 feeds, 2,994 rows per day, daily since 2026-11-03
- sources: World Pulse Daily Census (Lilambd) (https://huggingface.co/datasets/Lilambd/world-pulse-daily); arXiv API (https://arxiv.org/help/api); Crossref REST API (https://api.crossref.org); Zenodo REST API (https://developers.zenodo.org); Wikimedia Action and Pageviews APIs (https://wikimedia.org/api/rest_v1/); ECB euro reference rates (https://www.ecb.europa.eu/stats/policy_and_exchange_rates/euro_reference_exchange_rates/html/index.en.html); DefiLlama stablecoins API (https://stablecoins.llama.fi); GitHub REST search API (https://docs.github.com/en/rest/search); npm downloads API and pypistats.org (https://pypistats.org); Hugging Face Daily Papers (https://huggingface.co/papers); Stack Exchange API (CC BY-SA) (https://api.stackexchange.com); crates.io API (https://crates.io/data-access)
- known limits: History starts 2026-11-03, so changes appear from the second day; App Store, Hacker News and Docker Hub feeds excluded; GitHub top repositories excluded (owner usernames), counts only; a missing day means the census did not run

What moved in tech and money today, from public catalogues that only ever show *today*. Lilambd writes them down every day, so they become a history.

- **13 feeds**, 2,994 rows on the first day, loaded every day since 2026-11-03 (1 day so far; the API grows by one day every day)
- **Briefing**: one finished answer per day - headlines plus GitHub new repositories by language (341,057 on 2026-11-02), arXiv AI/ML submissions, most upvoted Hugging Face papers, AI SDK downloads with risers and fallers, Stack Overflow AI tags, crates.io, Crossref and Zenodo output, Wikipedia most read and growth, ECB rates (EUR/USD, USD/JPY...) and USD stablecoin supply ($297.5 bn) with the biggest coin moves
- **Today**: every feed's latest day with its top items and change against the previous day
- **Feeds**: any feed for any day (value, previous value, change and change %), and the history of one item (a package, a currency, a language edition, a tag)
- **Changes**: what is new or moved since a given day across all feeds (new papers, new top articles, new crates, rising downloads)
- **Search** across all feeds and days; **FX** with any base currency and conversion

Feeds: AI/agent SDK downloads (59 npm and PyPI packages), arXiv submissions in 54 categories, crates.io, Crossref DOIs by type/publisher/licence, ECB euro reference rates (29 currencies), GitHub new repositories (counts by language only), Hugging Face Daily Papers, stablecoin supply (80 coins, 209 chains), Stack Overflow tag counts, Wikipedia size (52 languages) and top 100 most read per language, Zenodo records.

Rows carry the day they were fetched and the date the numbers refer to (arXiv d-3, Wikipedia top d-2, counts of 'yesterday' d-1). A missing day means the census did not run, not zero. Not included: Apple App Store charts, Hacker News front page and Docker Hub pulls (terms unclear for resale), GitHub top repository lists (owner usernames). Sources: arXiv, Crossref, Zenodo (CC0 metadata); Wikimedia (CC0 statistics); European Central Bank; DefiLlama; GitHub; npm and pypistats.org; Hugging Face; Stack Exchange (CC BY-SA); crates.io. Census by Lilambd, CC BY-SA 4.0.

Routes:
- GET /pulse/fx: Exchange rates Params: base (Base currency (default EUR)), to (Target currency), amount (Amount to convert (with to)), date (YYYY-MM-DD, default latest)
- GET /pulse/days: Days loaded Params: limit (Days (max 400))
- GET /pulse/feeds: List feeds
- GET /pulse/today: Today across feeds Params: date (YYYY-MM-DD, default latest)
- GET /pulse/search: Search all feeds Params: q* (Text (2+ characters)), feed (Only one feed), from (First day), to (Last day), history (true = every day's row instead of one row per item), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /pulse/changes: Changes since a day Params: since (YYYY-MM-DD, default the day before the latest), date (Compare up to this day (default latest)), feed (Only one feed (ai_sdk_downloads | arxiv | crates_io | crossref | fx_ecb | github), kind (Only one kind within a feed (e.g. en, coin, language, npm)), type (all | new | moved), min_change_pct (Only moves of at least this many percent), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /pulse/briefing: Daily briefing Params: date (YYYY-MM-DD, default latest loaded day)
- GET /pulse/feeds/{feed}: One feed on one day Params: feed* (ai_sdk_downloads | arxiv | crates_io | crossref | fx_ecb | github_new_repos | hf), date (YYYY-MM-DD, default latest), kind (Kind within the feed (e.g. en, coin, chain, npm, pypi, ai, top100, type-name)), q (Text in name or key), sort (value | change | change_pct | rank | name), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /pulse/feeds/{feed}/history: History of one item Params: feed* (ai_sdk_downloads | arxiv | crates_io | crossref | fx_ecb | github_new_repos | hf), key (Exact key from /feeds/{feed}), q (Or text in the name), kind (Kind), from (First day), to (Last day), limit (Max rows (1-200, default 50)), offset (Rows to skip)

## States and Provinces of the World + Land Stats

- id / MCP tool: regions
- base: https://api.lilambd.com/regions
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/regions/search?sample=1
- data as of: not recorded; refresh: static; licence: Public domain (boundaries); CC BY 4.0 (indicators)
- coverage: 4,596 regions; land indicators 2000-2020 and forest loss since 2001
- sources: Natural Earth admin-1 (https://www.naturalearthdata.com/); Hansen GFC v1.12, GLAD GLCLU v2, JRC GSW v1.4, JRC GHSL R2023A, SoilGrids 2.0 (https://glad.umd.edu/)
- known limits: Population growth figures can be extreme where the 2000 baseline is small.

Every first-level administrative region on Earth, ready for dropdowns, maps, games and analytics.

- **4,596 regions** with ISO 3166-2 code, English, local and Japanese names, Wikidata id, area
- Satellite-derived land statistics per region: built-up share 2020, cropland share and change 2000-2020, forest cover 2000/2020, forest loss since 2001, surface water gained/lost 1984-2021, population growth 2000-2020, built-up growth 1975-2020, soil organic carbon
- Filter by country, search by any name, sort by area / forest loss / built-up / population growth

Sources: Natural Earth (public domain); Hansen GFC, GLAD GLCLU, JRC Global Surface Water, JRC GHSL, SoilGrids (CC BY 4.0). Computed by Lilambd.

Routes:
- GET /regions/search: Search regions Params: q (Name in any language), country (ISO 3166-1 alpha-2), sort (area | forest_loss | built | population_growth), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /regions/{iso3166_2}: Get a region Params: iso3166_2*

## What Is Rising in the World Today

- id / MCP tool: rising
- base: https://api.lilambd.com/rising
- price: $0.05 USDC per call
- free sample: https://api.lilambd.com/rising/days?sample=1
- data as of: not recorded; refresh: daily (when world-signals updates); licence: Derived counts; cite GDELT Project; npm/HF/GitHub public metadata
- coverage: one computed day per world-signals day; compares with the previous measured day
- sources: world-signals (Lilambd) (https://huggingface.co/datasets/Lilambd/world-signals); GDELT 2.0 (https://www.gdeltproject.org/); npm registry download counts (https://github.com/npm/registry/blob/main/docs/download-counts.md); Hugging Face Hub (https://huggingface.co); GitHub search (https://github.com)
- known limits: Gaps in world-signals (e.g. 10/22-10/31) mean the first day after a gap compares across the gap. Google Trends, Apple Podcasts and Steam parts of world-signals are excluded here.

A daily answer, not raw data: what jumped since the previous measured day across open software and world events.

- New npm packages and x3+ weekly-download jumps, new Hugging Face trending models/datasets/spaces, the week's fastest-starred new GitHub repos, and country x event-type spikes in GDELT (x3+, 50+ events)
- Grouped into themes (agents/MCP, OCR, 3D, speech, video, local LLM, security...) with a score = tables x items
- Every row carries the measured before/now values and the two days compared; no guesses
- /history?since=YYYY-MM-DD returns only the days after your last visit

Routes:
- GET /rising/days [free]: Free: list of computed days
- GET /rising/today: Latest day: themes and rising items Params: table, theme, limit
- GET /rising/{date}: One day (YYYY-MM-DD) Params: date*
- GET /rising/history: Every computed day after since (change feed) Params: since*, table, theme

## Which new rules on my topic are coming in Japan or the US, and when can I still comment?

- id / MCP tool: rules-ahead
- base: https://api.lilambd.com/rules-ahead
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/rules-ahead/answer?topic=%E5%BB%BA%E7%AF%89&sample=1
- data as of: 2026-11-03; refresh: Daily (rule_signals and pubcom_results loaded daily; e-Gov law revisions as loaded); licence: Japan: Public Data License / Government of Japan Standard Terms of Use 2.0 (CC BY 4.0 compatible), attribution e-Gov. US: Federal Register content is US public domain. Computed fields by Lilambd.
- coverage: JP: e-Gov public-comment drafts currently open (26 on 2026-11-03), recent e-Gov results with number of comments (33, results 2026-10-30..2026-11-02), promulgated-but-not-in-force revisions of the JP laws tracked in api_pub.egov_revisions (12 revisions, 7 laws, mainly pharma, narcotics, aviation, animal health). US: Federal Register proposed and final rules published 2026-08-04..2026-11-02 (1,147 documents) plus some older significant rules. 41 bilingual topics plus free-text matching.
- sources: e-Gov Public Comment (Japan, Digital Agency / Ministry of Internal Affairs and Communications) (https://public-comment.e-gov.go.jp/); e-Gov Law Search API (Japan) (https://laws.e-gov.go.jp/); US Federal Register API (Office of the Federal Register, NARA) (https://www.federalregister.gov/developers/documentation/api/v1)
- known limits: Topic matching is keyword based on title and category (plus agency for US) and can miss or over-include items. JP drafts table holds only currently open drafts, not closed history; JP titles have no English translation. JP law amendments cover only laws tracked in egov_revisions, not all of Japanese law; e-Gov placeholder in-force dates (2117-12-31) are shown as not yet fixed. US coverage starts 2026-08. Routine US items hidden by default. Not legal advice.

One finished brief per topic: Japanese e-Gov public-comment drafts still open, US Federal Register proposed rules with open comment windows, US final rules and Japanese law amendments published but not yet in force, and recent Japanese comment results with the number of comments received. Plain English or Japanese keywords (drone, AI, food labeling, tax, 建築, 医薬品). Deadlines, days left, counts and topic matching computed by Lilambd from official registers; every item links to its official page. Japanese titles stay in Japanese; US items carry title_en. Facts only, not legal advice.

Routes:
- GET /rules-ahead/answer: Finished brief for one topic: open comment windows with days left, rules not yet in force, recent JP results with comment counts, top agencies, links Params: topic* (Plain English or Japanese keyword, e.g. drone, AI, food labeling, tax, labor, 建築), country (JP or US; omit for both), limit (Items per section (default 10, max 50)), routine (include to keep routine US items (airworthiness directives, safety zones, drawbr)
- GET /rules-ahead/topics: Which topics are covered, with live counts of open comment windows and items not yet in force per country Params: country (JP or US; omit for both), limit (Default 50, max 200)
- GET /rules-ahead/upcoming: List of coming rules: open comment windows and items not yet in force, filterable by country, kind, agency, category, topic and dates Params: country (JP or US), status (ahead (default: open comment windows + not yet in force), open, effective_ahead,), topic (Same keyword matching as /answer), kind (rule, significant_rule, guideline, ministerial_ordinance, notice, other, law_ame), stage (proposal, final, promulgated, result), agency (Substring of the agency name (English for US, Japanese for JP, e.g. 国土交通省)), category (Substring of the category (US CFR topics; JP e-Gov categories such as 建築、住宅)), since (Published on or after (YYYY-MM-DD)), until (Key date (comment deadline or in-force date) on or before (YYYY-MM-DD)), sort (date (key date ascending, default) or newest), limit (Default 50, max 200), offset (Rows to skip)

## World Science Output by Country and Field

- id / MCP tool: science
- base: https://api.lilambd.com/science
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/science/gaps?sample=1
- data as of: 2026-10-14; refresh: monthly (rebuild script in HF Lilambd/world-science); licence: CC0 (OpenAlex)
- coverage: all 464,765,485 OpenAlex works: 18,007 country-years, 3,429 field-years, 8,716 rising subfield-country pairs, 252 subfields
- sources: OpenAlex works snapshot 2026-10-14 (https://openalex.org); Natural Earth (population) (https://www.naturalearthdata.com)
- known limits: Compare papers only: dataset records are inflated by mass registration (about 18M 2025 datasets carry a Japanese affiliation). XX = no affiliation country (4.3M papers in 2025). Records added before 2025-11 cannot be dated by month.

The shape of world research in one API, counted from every record in OpenAlex (snapshot 2026-10-14, 464,765,485 works).

- **Countries x years**: papers (whole and fractional counting), preprints, datasets, open-access papers, citations
- **Rank** countries for a year by papers, papers per million people, open-access share or citations
- **Gaps**: countries with the fewest papers per million people, so thin places are seen
- **Fields x years** for 26 fields, with growth against five years before
- **Rising**: subfield x country pairs where 2025 output jumped against the 2020-2024 average (e.g. Health Informatics in Japan: index 292)
- 252 subfields with field and domain

Only 'papers' are compared (article, review, book, letter, editorial, dissertation); dataset records are counted apart because some repositories mass-register them. Country XX = no affiliation country. Source: OpenAlex (CC0); population from Natural Earth.

Routes:
- GET /science/gaps: Gaps Params: year (Year), min_population (Minimum population), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /science/rank: Rank countries Params: year (Year), by (papers | per_million | open_access_share | citations), min_population (Minimum population), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /science/fields: Fields in a year Params: year (Year), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /science/rising: Rising subfields Params: country (ISO 3166-1 alpha-2), subfield (Subfield id), field (Field name text), min_papers (Minimum papers in 2025), sort (index | papers), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /science/subfields: Subfields Params: q (Text), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /science/countries/{iso2}: Country history Params: iso2*, from (First year), to (Last year), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /science/fields/{field_id}: Field history Params: field_id*, from (First year), to (Last year), limit (Max rows (1-200, default 50)), offset (Rows to skip)

## Security ledgers: exploited CVEs (CISA KEV + EPSS + NVD), breach disclosures, change feed

- id / MCP tool: security
- base: https://api.lilambd.com/security
- price: $0.01 USDC per call
- free sample: https://api.lilambd.com/security/hot?sample=1
- data as of: 2026-11-01; refresh: KEV every 4h, NVD every 2h, EPSS daily, HIBP daily; licence: per source, see sources; attribution in every response
- coverage: KEV 1,766 exploited CVEs; EPSS 398,784 CVEs; NVD 417,673 CVEs (full history 1988-, refreshed every 2h); breaches: HIBP 934 + Japan disclosures 55; incidents ~7,200 (SEC 8-K Item 1.05, Washington and California AG notices, CISA/JPCERT/IPA advisories)
- sources: CISA Known Exploited Vulnerabilities (https://www.cisa.gov/known-exploited-vulnerabilities-catalog); FIRST EPSS (https://www.first.org/epss); NVD CVE API 2.0 (https://nvd.nist.gov/developers/vulnerabilities); Have I Been Pwned breaches (https://haveibeenpwned.com/api/v3/breaches); Lilambd JP disclosure ledger v0.2 (https://lilambd.com)
- known limits: Organisation-level only, no personal records. No phishing URL feed yet (licences). JVN offline until late Nov 2026. Vendor matching by KEV names and NVD CPE strings.

Fresh, sourced security facts in one joined shape. Every actively exploited CVE in CISA KEV joined with its EPSS score (FIRST, refreshed daily) and NVD CVSS/CWE/CPE; NVD CVEs (418k, refreshed every 2 hours); organisation-level breach records from Have I Been Pwned plus a Japan disclosure ledger built from company announcements; and a since= change feed of what was added or moved. Every row has a source URL; every response has data_as_of per source. No personal records. Plus an incident ledger: SEC 8-K Item 1.05 cyber-incident filings, Washington and California Attorney General breach notices (7,000+ since 2012, with cause, attack type, residents affected), and CISA / JPCERT / IPA advisories, refreshed daily.

Routes:
- GET /security/hot: Recent CVEs not yet in KEV, ranked by EPSS (likely to be exploited next) Params: days, vendor (NVD CPE vendor, e.g. microsoft), limit
- GET /security/kev: Actively exploited CVEs (CISA KEV) with EPSS and CVSS joined, newest first Params: vendor, product, since (date_added after YYYY-MM-DD), ransomware (known = only CVEs CISA links to ransomware campaigns), q, limit, offset
- GET /security/gaps [free]: Free. What is missing or thin, and why
- GET /security/stats [free]: Free. Row counts and last load per source
- GET /security/changes: Change feed: new KEV entries, EPSS jumps, new/modified CVEs, new breaches Params: since* (ISO-8601 time), after_seq, face, limit
- GET /security/vendors: Vendors ranked by actively exploited CVEs, ransomware-linked count, added in last 90 days Params: q, limit
- GET /security/breaches: Organisation-level breach records (HIBP + Japan disclosures) by domain, name, country or date Params: domain, q, country, since (disclosed after YYYY-MM-DD), limit
- GET /security/cve/{id}: One CVE: exploited in the wild?, KEV entry, NVD record, EPSS, change history Params: id*
- GET /security/incidents: Security incidents, events and cases in one table: SEC 8-K Item 1.05 filings, Washington and California AG breach notices, CISA / JPCERT / IPA advisories, HIBP and Japan breach records. Organisations only. Params: q (organisation name contains), country, state, kind, source, attack_type, cve, since (disclosed after YYYY-MM-DD), until, limit, offset
- GET /security/incidents/trends: Regulator-reported incidents per month by source and attack type, plus top sectors Params: months

## Lilambd World Signals Daily Change by Country

- id / MCP tool: signals
- base: https://api.lilambd.com/signals
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/signals/days?sample=1
- data as of: 2026-11-03; refresh: daily: load the new data/YYYY-MM-DD/gdelt_events.csv from the Hugging Face shelf after its morning (JST) build; licence: GDELT: free for any use with attribution; Lilambd/world-signals compilation CC BY 4.0
- coverage: 19,185 country x CAMEO-root rows, 126,256 GDELT events, 228 GDELT (FIPS 10-4) country codes, 31 collected days 2026-09-22 to 2026-11-03; 20 CAMEO roots; 111 FIPS codes mapped to ISO2
- sources: Lilambd World Signals (Hugging Face) (https://huggingface.co/datasets/Lilambd/world-signals); GDELT 2.0 Event database (The GDELT Project) (https://www.gdeltproject.org/)
- known limits: Each day is about one hour of GDELT export, not 24 hours; counts are news coverage, not verified incidents. Days not collected: 2026-10-15, 2026-10-16, 2026-10-22 to 2026-10-31 (never back-filled). GDELT FIPS codes: 111 mapped to ISO2, the rest listed under /gaps. No admin1, no coordinates. Google Trends, Apple podcasts, Steam, GitHub, HF trending and package downloads from the same shelf are excluded (terms unclear, personal names, or overlap with agents).

What changed today in the world's news events, per country. Daily GDELT 2.0 event counts by country and CAMEO root code (20 event types from public statement to mass violence) with mean Goldstein score, collected every morning since 2026-09-22: 19,185 country x event-type rows covering 126,256 events in 228 GDELT country codes over 31 collected days (2026-09-22 to 2026-11-03; 12 days not collected). One call to /today/{iso2} returns today's totals for a country against the previous collected day and the 7-day baseline, its rank, conflict share and the biggest changes by event type. Also: rankings, rising country x event pairs, time series per country. Source: Lilambd/world-signals (CC BY 4.0), GDELT Project (free with attribution).

Routes:
- GET /signals/days: Collected days with totals
- GET /signals/gaps: Missing days and unmapped country codes
- GET /signals/rank: Countries ranked for one day Params: by (events, conflict_share or goldstein), min_events (minimum events), date (YYYY-MM-DD; default latest collected day), limit (rows per page, 1-200 (default 50)), offset (rows to skip)
- GET /signals/cameo: The 20 CAMEO root codes
- GET /signals/events: Rows for one day Params: country (ISO2 or FIPS), cameo (CAMEO root 01-20), date (YYYY-MM-DD; default latest collected day), limit (rows per page, 1-200 (default 50)), offset (rows to skip)
- GET /signals/rising: Country x event type rises vs 7-day baseline Params: min_events (minimum events today), date (YYYY-MM-DD; default latest collected day), limit (rows per page, 1-200 (default 50)), offset (rows to skip)
- GET /signals/countries: FIPS to ISO2 keys
- GET /signals/cameo/{root}: Countries for one event type Params: root* (CAMEO root 01-20), date (YYYY-MM-DD; default latest collected day), limit (rows per page, 1-200 (default 50)), offset (rows to skip)
- GET /signals/today/{iso2}: What changed today for country X Params: iso2* (ISO 3166-1 alpha-2 (or GDELT FIPS code)), date (YYYY-MM-DD; default latest collected day)
- GET /signals/country/{iso2}: Daily totals for one country Params: iso2* (ISO2 or FIPS), days (days back, max 120), date (YYYY-MM-DD; default latest collected day), limit (rows per page, 1-200 (default 50)), offset (rows to skip)

## World Music Instruments and Singing Styles

- id / MCP tool: sound
- base: https://api.lilambd.com/sound
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/sound/names?sample=1
- data as of: 2026-11-02; refresh: static snapshot; licence: The Global Jukebox and D-PLACE CCMC: CC BY 4.0; MIMO: CC0; compilation CC BY 4.0 Lilambd
- coverage: 1,466 societies; 1,780 instruments; 643 Hornbostel-Sachs classes; 2,724 instrument names in 12 languages; 776 ensembles; 5,776 songs; 700 genres; 1,575 sound sources; 1,007 song functions; 37 Cantometrics axes (216 codes) with 37,845 society values
- sources: The Global Jukebox (https://github.com/theglobaljukebox); D-PLACE CCMC (Natural History of Song 2) (https://github.com/D-PLACE/dplace-cldf); MIMO Musical Instrument Museums Online (https://github.com/modavis/mimo)
- known limits: Coverage follows the research archives, so some regions have few societies. Performer and recorder names are left out. Scala tunings and secondary lists are excluded (licence unclear). Society coordinates are archive centroids. /styles needs society_id or var_id.

Cross-cultural music data for music apps, games, education, sound design and AI agents, built from open research archives.

- **Societies**: 1,466 societies with coordinates, region and counts; full profile per society (instruments, songs, singing style, song functions, ensembles); **nearby** societies for any coordinate
- **Instruments**: 1,780 instrument records by society with local names, Hornbostel-Sachs number and class (idiophone, membranophone, chordophone, aerophone)
- **Classes**: 643 Hornbostel-Sachs classes with descriptions; **names**: 2,724 instruments in English plus German, French, Spanish, Italian, Dutch, Swedish, Polish, Catalan, Basque, Chinese and Korean
- **Songs**: 5,776 coded songs with genre, sound sources, vocalist gender, year and duration; genre and sound-source summaries
- **Singing style**: 37 Cantometrics axes (group organisation, melody, rhythm, voice quality...) with each society's typical value
- **Song functions**: dance, lullaby, healing, love, mourning... by society

Performer and recorder names are left out. Sources: The Global Jukebox (CC BY 4.0); D-PLACE CCMC / Natural History of Song 2 (CC BY 4.0); MIMO classes and names (CC0).

Routes:
- GET /sound/names: Instrument names in 12 languages Params: q (Name in any of the languages), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/songs: Songs Params: q (Song or society), genre (Genre text), sound_source (Voice or instrument text), region (Region), society_id (Society id), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/genres: Genres Params: q (Text), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/styles: Singing style values Params: society_id (Society id), var_id (Axis id), code (Code), region (Region), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/classes: Hornbostel-Sachs classes Params: q (Text or code), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/ensembles: Ensembles Params: q (Instrument text), region (Region), country (Country), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/functions: Song functions Params: function (e.g. Dance, Lullaby, Healing), region (Region), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/societies: Societies Params: q (Society name), region (Region text), country (Country text), lat (Latitude), lon (Longitude), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/instruments: Instruments Params: q (Name or local name), hs_class (idiophone | membranophone | chordophone | aerophone), hs_number (Hornbostel-Sachs prefix), region (Region text), country (Country text), society_id (Society id), lat (Latitude), lon (Longitude), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/styles/axes: Singing style axes Params: q (Text), var_id (e.g. line_1), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/sound-sources: Sound sources Params: q (Text), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /sound/instruments/{id}: Get an instrument Params: id*
- GET /sound/classes/{hs_code}: Get a class Params: hs_code*
- GET /sound/societies/{society_id}: Society profile Params: society_id*

## Travel meds: can I bring this medicine into this country?

- id / MCP tool: travel-meds
- base: https://api.lilambd.com/travel-meds
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/travel-meds/answer?country=JP&sample=1
- data as of: 2026-11-01; refresh: weekly recheck of source pages; licence: facts from public government pages; Lilambd shaping CC BY 4.0
- coverage: 9 countries (JP, SG, GB, US, TH, CN, KR, DE/Schengen, AE), 19 rows
- sources: Government pages per country (MHLW, HSA, GOV.UK, FDA, Thai FDA/embassy, Shanghai gov, Korea MFDS via MOFA, BfArM)
- known limits: v0: 9 countries. AE row is from a 2018 news article citing MOHAP (official page blocks automated reading). Australia not yet (TGA blocks automated reading). Not medical or legal advice; rules change, check the cited page before travel.

Official rules for travellers carrying personal medicines across borders: how many days of supply, which classes need a permit, how and how early to apply, documents, and what is banned outright. Every row cites the government page it was read from and when. Facts only, not medical or legal advice.

Routes:
- GET /travel-meds/answer: One finished answer for a trip: rules for a country and drug class, with the steps to take Params: country*, class (prescription, otc, psychotropic, narcotic, controlled), days
- GET /travel-meds/country: All rows for one country Params: country*
- GET /travel-meds/countries [free]: Which countries are covered

## Unit Converter and Physical Constants

- id / MCP tool: units
- base: https://api.lilambd.com/units
- price: $0.005 USDC per call
- free sample: https://api.lilambd.com/units/search?sample=1
- data as of: 2024-05-09; refresh: static; licence: CC BY 4.0 (QUDT); US government work (NIST)
- coverage: 2,932 units, 355 constants (CODATA 2022), 3,352 isotopes
- sources: QUDT (https://qudt.org); CODATA 2022 / NIST (https://physics.nist.gov/cuu/Constants/)
- known limits: QUDT version not recorded; the date is the CODATA 2022 release.

One clean JSON API for apps, games, calculators and AI agents that need correct unit maths.

- **2,932 units** from QUDT with SI multipliers and offsets (temperature handled correctly)
- Type-safe conversion: incompatible units are rejected with the quantity kinds shown
- **355 physical constants** (CODATA 2022) with uncertainty and exact flag
- **3,352 isotopes** with relative atomic mass (NIST)
- Accepts ids, symbols, UCUM codes and plain names (`mi`, `km`, `degF`, `[lb_av]`, `kelvin`)

Sources: QUDT (CC BY 4.0), CODATA 2022 / NIST SRD. Attribution is included in every response.

Routes:
- GET /units/search: Search units Params: q (Text to search), quantity_kind (e.g. length, mass, pressure), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /units/{unit}: Get a unit Params: unit*
- GET /units/convert: Convert a value Params: value* (Number to convert), from* (Source unit (id, symbol, UCUM or name)), to* (Target unit)
- GET /units/isotopes: List isotopes Params: element (Element symbol), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /units/constants: Search physical constants Params: q (Text to search), limit (Max rows (1-200, default 50)), offset (Rows to skip)
- GET /units/constants/{id}: Get a constant Params: id*
- GET /units/quantity-kinds: List quantity kinds

## World Dishes: national and regional dishes by country

- id / MCP tool: world-dishes
- base: https://api.lilambd.com/world-dishes
- price: $0.02 USDC per call
- free sample: https://api.lilambd.com/world-dishes/dish?q=satay&sample=1
- data as of: not recorded; refresh: Snapshot of 2026-11-03; yearly refresh with the world food ledger; licence: Wikidata CC0 1.0
- sources: Wikidata SPARQL (dish -> country of origin / cuisine), fetched 2026-11-03 (https://query.wikidata.org/)

Look up the dishes Wikidata links to a country (ISO3), or search a dish by name in English or Japanese. Returns Wikidata QID, English and Japanese names and every country the dish is linked to. Values as recorded in Wikidata; nothing estimated. Coverage reflects Wikidata, so some countries have few or no dishes.

Routes:
- GET /world-dishes/dish: Search dishes by name Params: q*, limit
- GET /world-dishes/dishes: Dishes of one country Params: iso3*, limit

## How does this country's education system measure up?

- id / MCP tool: world-education
- base: https://api.lilambd.com/world-education
- price: $0.05 USDC per call
- free sample: https://api.lilambd.com/world-education/answer?country=JP&sample=1
- data as of: 2026-11-03; refresh: Monthly rebuild from the WDI mirror and the OpenAlex snapshot; licence: World Bank WDI CC BY 4.0 (UNESCO UIS series); OpenAlex CC0; ROR CC0. Computed fields by Lilambd, reuse with attribution.
- coverage: 209 countries, 14 headline indicators, latest year per indicator (some years are old: always read year)
- sources: World Bank World Development Indicators (UNESCO Institute for Statistics for SE.* series) (https://github.com/open-numbers/ddf--open_numbers--world_development_indicators); OpenAlex institutions snapshot 2026-10-14 (https://openalex.org); Research Organization Registry (https://ror.org)
- known limits: Years differ by country; gross enrolment can exceed 100%; some series have breaks (e.g. Zimbabwe education spend); institution counts include all OpenAlex institution types.

Latest measured education indicators for 209 countries (literacy, enrolment at each level, completion, spending, pupils per teacher, out-of-school children, R&D spend, researchers, tertiary gender parity), each with its year and World Bank series code, plus the number of research and education institutions in the country (OpenAlex, all with ROR ids). Compare up to 6 countries or rank all countries on one indicator. Facts only; no personal data.

Routes:
- GET /world-education/rank: All countries ranked on one indicator Params: indicator, since, limit
- GET /world-education/answer: One country brief Params: country*
- GET /world-education/compare: Up to 6 countries side by side Params: countries*
- GET /world-education/coverage: Countries and indicators covered

## World Power: country electricity profile

- id / MCP tool: world-power
- base: https://api.lilambd.com/world-power
- price: $0.5 USDC per call
- free sample: https://api.lilambd.com/world-power/answer?country=JP&sample=1
- data as of: 2026-11-03; refresh: yearly; monthly Ember load prepared; licence: CC BY 4.0 (attribution in every response)
- coverage: 214 countries + World and regions; 34,936 plants in 167 countries
- sources: Ember Yearly Electricity Data (2026) (https://ember-energy.org/data/yearly-electricity-data/); Energy Institute Statistical Review of World Energy (2025) (https://www.energyinst.org/statistical-review); Our World in Data energy-data (https://github.com/owid/energy-data); WRI Global Power Plant Database v1.3.0 (https://datasets.wri.org/dataset/globalpowerplantdatabase)
- known limits: Plant capacity is a 2017-2021 snapshot and misses newer plants; null = not recorded, never zero; about half of the countries have 2024 as their latest year

One call returns a country's electricity in numbers: generation by fuel (TWh) and shares for 2025 (or 2024), change vs the previous year, fossil and low-carbon share, carbon intensity (gCO2e/kWh), demand, net imports, power-sector emissions, installed plant capacity by fuel and the five largest plants. 214 countries plus World and regions. Values as published by Ember, the Energy Institute and WRI; nothing estimated.

Routes:
- GET /world-power/: Discovery: routes and coverage
- GET /world-power/mix: Generation by fuel for one country-year Params: country*, year
- GET /world-power/answer: Country electricity profile Params: country* (ISO 3166-1 alpha-2 or alpha-3, or World), year
- GET /world-power/countries: All areas with latest generation and carbon intensity Params: limit, offset