Portal and platform-level changes. Newest first.2026-11-06#
Sugra Entity, POST /api/v1/entity/screen and POST /api/v1/entity/rf-watchlists/screen: an answer now says how many records matched. screening.total_matches counts them, screening.total_matches_exact is false when that count is a lower bound, and screening.truncated is true when the answer may have left some out; the new optional max_matches (1 to 100, default 10) widens it. The window is chosen by name score, before dob, country and nationality are applied, and matches with an equal name score now enter it one per list in turn, with matches flagged incomplete_coverage still after the others at the same name score, so the lists whose names sort last are no longer cut by the records of the lists before them. dob, country and nationality then re-score the matches in the window and can change their order, so matches with an equal final score are not guaranteed to alternate between lists. This can change which ten records a screen of a common name returns, and their order. Because status is decided over the matches in the window, it can change as well when records with an equal name score differ in their dob, country or nationality evidence, or in whether they can be a hit at all (a PEP match or one flagged incomplete_coverage never is); at an equal name score a match flagged incomplete_coverage enters the window after the others, so a window too short for all of them leaves it out before any other match of that score. The batch screen and the composed envelope's screening carry the three fields with their window of 10.2026-11-05#
Sugra MCP is listed in the GitHub MCP Registry for GitHub Copilot and VS Code: View on GitHubThe listing's Install MCP server button adds the hosted server to VS Code, VS Code Insiders or the GitHub Copilot app; MCP covers every install path. Sugra MCP is now in all three: Anthropic's Connectors Directory, OpenAI's Plugins Directory and the GitHub MCP Registry.
Sugra Research, official statistics: the whole world, continents and world regions (groups of countries such as Sub-Saharan Africa) are places now. place=world takes the whole world, /places/resolve gives the code of a continent or a world region by its name, and /places/{indicator}?parent=world lists countries, continents or world regions.Sugra API Skills: four skills for the data itself, markets-data, macro-data, news-data and earth-data, name the call for each question in their area. Every plugin carries all ten, and Ecosystem lists them.2026-11-04#
Sugra Research, official statistics: beyond the 81 indicators picked by hand, /statistics/search now finds the narrower series of the catalog, most of them a cut of a broader measure and most of them for the United States. The 2026-11-01 release served only figures an agency published itself. Where an agency published none, an answer can now be a figure computed from its data (a sum, count or average) or a model projection, which may lie in the future, and meta.notes says which: "Computed from the agency's data: a sum, count or average the agency does not publish itself." or "A model projection, not an observed value."2026-11-01#
Sugra Research: official statistics by place under /api/v1/statistics, 81 indicators (population, income, poverty, unemployment, housing, health, education, crime, energy, air quality) for places from countries down to US counties, ZIP code areas and European NUTS regions. /search?q= finds an indicator by meaning and /indicators/{indicator} describes one; /series/{indicator}?place= returns its series for one place; /places/{indicator}?parent=&type= lists every place of a type inside a region; /places/resolve turns a place name or coordinates into a code. Places take open codes (ISO 3166, FIPS, NUTS, ZIP code areas), data.source names the agency that published the values, and projections and model estimates are never served.
/api/v1/fred/series/{series_id}: new optional as_of (YYYY-MM-DD) returns the observations as they were published on that date, before later revisions. The response adds as_of and vintage_date; the title, units, frequency, seasonal adjustment, last_updated, observation_start and observation_end are those of that date (a series rebased since keeps the units its values were published with, and observation_end is the last observation that existed then). A malformed date, a date before 1776-07-04, a date after today in US Central time and a date before the series' first release answer 400. Without as_of nothing changes.
New GET /api/v1/fred/series/{series_id}/vintages lists the dates a FRED series was released or revised, newest first, bounded by limit.
New names on these guides: Sugra MCP connects an AI assistant or agent, Sugra API serves your code by key, and both run on Sugra Data, the data service whose seven directions the guides describe. The MCP server is now named Sugra MCP. Addresses, page links, endpoints and keys do not change.
The MCP page has a Plans section for Free and Personal, and the cost of each Sugra MCP tool moved there from Request cost. 2026-10-08#
/api/v1/fred/series/{series_id}: with a units transform other than lin, units and units_short now name the transform (for example Percent Change from Year Ago) and the series' own labels move to the new level_units and level_units_short. observation_start and observation_end now give the dates of the series' first and latest observations; they used to repeat the requested window, or 1600-01-01 and 9999-12-31 when none was asked for.
meta.data_time now names the newest row on /api/v1/commodities/energy/petroleum, /api/v1/commodities/energy/natural-gas, /api/v1/commodities/energy/petroleum-stocks, /api/v1/commodities/{commodity_id} and /api/v2/futures/{root}/historical. These showed the time of the response, and the commodity history its oldest row. On /api/v1/weather/us/forecast and /api/v1/weather/us/forecast/hourly it is the time the forecast was last updated, which updated now carries too. /api/v1/commodities/prices, /api/v1/commodities/energy/steo and /api/v1/commodities/agriculture/grains no longer report a meta.data_age_days of 0; each row keeps its own date, or year for grains.
/api/v1/commodities/energy/petroleum: limit counts per benchmark, so Brent and WTI each get up to that many daily prices.
2026-10-02#
Sugra Macro: us/pce now serves the PCE price index (it returned nominal consumer spending, which moves to its own key, us/nominal-pce); on weekly series wow_bps compares with the previous week and year-over-year changes are filled in last_n requests; JOLTS rates, two inventory-to-sales ratios, the financial stress indices and two monthly exchange rates take the transforms that fit their units, listed per key in /api/v1/macro/catalog.2026-09-24#
The Sugra API MCP server is published in Anthropic's Connectors Directory: Add to ClaudeThe directory serves Claude on the web, desktop and mobile, Claude Code and Cowork. Connect the Sugra API MCP server there and sign in with your Sugra account; MCP covers every install path. It is also listed in OpenAI's Plugins Directory for ChatGPT and Codex, since July 2026.
The Sugra API Skills are listed in OpenAI's Plugins Directory too; MCP has the Claude Code commands. Already added Sugra to Claude as a custom connector? Remove it before connecting from the directory, or you will have two connections.
The hosted MCP server now lives at https://mcp.sugra.ai, the server the directory listing connects to. Existing setups keep working unchanged; MCP has the connector URL for a manual setup. 2026-09-12#
Statistical agencies wave, merged from 29 August to 12 September:New national statistical offices: Statistics Netherlands (CBS) at /api/v1/cbs/{series}, plus any StatLine table at totals under /api/v1/cbs/table/{table_id}; the Swiss Federal Statistical Office (FSO) at /api/v1/bfs/{series}; Statistics Poland (GUS) at /api/v1/gus/{series}; and Italy's ISTAT at /api/v1/istat/{series}. Each lists its series under /catalog.
Destatis, the German Federal Statistical Office: browse the statistics catalogue at /api/v1/destatis/statistics, fetch any table by its GENESIS code at /api/v1/destatis/table/{code}, and queue an extraction too large for a direct request with POST /api/v1/destatis/extractions, then collect it at /api/v1/destatis/extractions/{extraction_id}.
INSEE adds its Melodi national datasets (population, wages, housing stock, poverty rate) to /api/v1/insee/{series}.
US Census Bureau: monthly trade by commodity, partner, state, port, end use and advanced technology under /api/v1/census/trade/...; County and ZIP Code Business Patterns, nonemployer statistics, business dynamics and the Annual Business Survey under /api/v1/census/business/...; economic indicators at /api/v1/census/indicators; population estimates and international demography under /api/v1/census/population/...; state and local public finance under /api/v1/census/public-finance/...; and small-area poverty, health insurance and community resilience under /api/v1/census/welfare/....
Eurostat PRODCOM sold production at /api/v1/eurostat/prodcom, and the open-data gateway adds the Irish, Swiss and German national portals (all portals are listed at /api/v1/opendata/portals).
2026-09-09#
Sugra Entity resolves French companies: siren and siret join lei and vat as anchors on /api/v1/entity/{anchor}/{value}, resolved through INSEE Sirene. Like vat, they carry no ownership graph.2026-09-05#
Sugra Finance: /api/v1/forex/rates serves a full currency reference panel built from central-bank fixings. Each row carries its fixing date, cadence and source.2026-09-02#
Maritime chokepoints, each from the state or intergovernmental record of the passage, merged from 31 August to 2 September:Strait of Hormuz: weekly full transits from UKMTO at /api/v1/maritime/chokepoints/hormuz/transits, Gulf crude exports from JODI at /api/v1/maritime/chokepoints/hormuz/gulf-exports, and a destination-side crude-import floor at /api/v1/maritime/chokepoints/hormuz/import-floor.
JMIC advisory transit counts for Hormuz and Bab el-Mandeb at /api/v1/maritime/chokepoints/jmic/advisory.
Strait of Malacca approaches: Singapore throughput from the MPA at /api/v1/maritime/chokepoints/malacca/throughput.
Panama Canal daily transits from the operator's own record at /api/v1/maritime/chokepoints/panama/transits.
Where a chokepoint count covers only the ships visible to AIS, the response now says so.
2026-08-31#
The response envelope states how old its data is: meta.data_age_days, present only when the source stated an exact time for the data. When it is absent, the age is unknown, not zero.2026-08-21#
Upstream failures are typed. Instead of a blanket 503, endpoints answer an upstream failure with an object that names the source and the kind of failure (upstream_error, upstream_timeout, upstream_unreachable and others), with a retry_hint that says whether to retry or what to change. Families adopted it one by one through September; Rate limits and errors lists the kinds.2026-08-10#
The volatility and options-statistics wave lands on Sugra Finance - six releases in one pass:Volatility futures: the VX root joins /api/v2/futures (curve with a contango flag, contract chain, per-contract daily history back to 2013) on a declared settlement basis (price_type: "settlement").
Exchange-wide put/call statistics: /api/v1/cboe/put-call-ratio (latest session, every published ratio plus per-product volume and open interest) and /put-call-ratio/history - a stitched daily series per family reaching back to 2006.
The volatility term structure gains the 1-day tenor: six points, 1d to 1y.
The full index catalog: /api/v1/cboe/indices/all - 2,481 indices, paged and filterable; /index/{symbol} now serves daily history for any of them.
Per-minute options flow: /api/v1/cboe/intraday/{symbol} - the latest session's one-minute bars with calls/puts/total options volume inside every minute.
Cash-settled index chains: /api/v1/cboe/options/_SPX|_NDX|_RUT documented and engineered at full size, with expiry=YYYY-MM-DD slicing and an always-included expirations list.
Data honesty rules apply throughout: placeholder zeros are served as null, absent series answer 404 (never a fabricated empty), and settlement-basis payloads declare their T-1 nature in-band.2026-08-09#
Macro and options analytics, merged from 5 to 9 August:Facet navigation over every FRED series: /api/v1/macro/facets narrows, /api/v1/macro/series selects.
The release calendar at /api/v1/macro/releases honors its filters; derived US net liquidity at /api/v1/macro/net-liquidity; a central bank meeting calendar at /api/v1/macro/cb-calendar.
A derived US macro regime at /api/v1/macro/regime.
Options helpers: /api/v1/options/{symbol}/strike-by-delta picks one contract by delta without downloading the chain, /api/v1/options/{symbol}/skew gives the 25-delta skew and ATM implied volatility for one expiry, and /api/v1/options/{symbol}/iv-rank ranks the current ATM implied volatility against its history.
2026-08-08#
A 401 on an /api/... path answers in JSON instead of an HTML help page. The body carries a stable code (missing_api_key, invalid_api_key or unauthorized), a reason and a hint; see Authentication.2026-07-30#
Transport on Sugra Earth, merged from 19 to 30 July:Weather-Disruption-Risk, an index Sugra derives, at a point or airport (/api/v2/transport/disruption-risk), across the forecast horizon (/api/v2/transport/disruption-risk/forecast) and along a corridor (/api/v2/transport/disruption-risk/route).
Observed US air-traffic delay programs from FAA NAS Status at /api/v2/transport/aviation/delays.
Port-call throughput deviation with a weather risk overlay, from IMF PortWatch, at /api/v2/transport/ports/congestion.
Port calls, schedule against actual, from Fintraffic Portnet at /api/v2/transport/ports/port-calls.
Road-disruption events in France from DATEX II (Bison Fute) at /api/v2/transport/road/events.
2026-07-25#
Sugra Entity screens a name or a Russian identifier against Russian regulatory watchlists: POST /api/v1/entity/rf-watchlists/screen, with the covered lists at /api/v1/entity/rf-watchlists/sources.2026-07-21#
The daily quota belongs to the account, not to the key.All keys of an account share one daily quota. Before, each key had its own, so adding keys raised the limit; it no longer does. Each key keeps its own usage counter.
A 429 over the quota names the daily limit, your current plan and a link to upgrade.
2026-07-18#
Autosync from CI: the portal is regenerated from the OpenAPI specification on every API deploy.
Guide pages: getting started, authentication, rate limits, and per-direction overviews.
Try-it environment: send authenticated requests directly from endpoint reference pages.
Weather v2 on Sugra Earth: an hourly and daily forecast up to 16 days ahead, worldwide, at /api/v2/weather/forecast, and history from 1940 at /api/v2/weather/history. /api/v1/weather/forecast and /api/v1/weather/history are deprecated in their favor.Endpoint changes#
Endpoint-level changes (new endpoints, parameters, response fields) ship continuously and are not itemized here. The OpenAPI specification at https://sugra.ai/openapi.json is always current; the portal syncs to it on every deploy.The spec is public and requires no API key. Modified at 2026-11-07 02:57:12