v1.1.0

agrobr now spans 40 public data sources for Brazilian agriculture behind one semantic layer. This release adds new
market-positioning, sugarcane, shipments and land-registry sources, ships full English documentation, and restores
several integrations after upstream changes.

New sources & data

• CFTC Commitments of Traders — weekly fund positioning across 12 agri futures (CBOT/CME/ICE), 2006+ —
  cftc.cot("soja")
• UNICA — Center-South sugarcane crush, sugar/ethanol mix and ATR (fortnightly PDF + historical XLSX) —
  unica.moagem_quinzenal("cana")
• ANEC — weekly port shipments by product — anec.embarques(ano=2026)
• INCRA Acervo Fundiário — certified rural parcels (SIGEF/SNCI) + agrarian settlements, with tabular and geo
  variants —
  acervo_fundiario.sigef_geo("MT")
• CEPEA Robusta/Conilon coffee — cepea.indicador("cafe_robusta")
• INMET full-year hourly history — no token required — inmet.historico("A701", 2025)
• ComexStat — pesticide imports (NCM 3808) via comexstat.importacao(produto="defensivos")

Documentation

• Full English documentation (~125 pages, PT + EN) with language switcher (mkdocs-static-i18n)
• Doc↔code audit fixing ~70 divergences (signatures, schemas, parameters)

Reliability (30+ source fixes)

• Integrations fixed after upstream changes: ANTAQ, IBAMA, B3 open interest, UN Comtrade, PRODES/deforestation,
  ICMBio
• Centralized HTTP retry re-enabled across CEPEA, IBGE/sidrapy, INMET, NASA POWER and CONAB
• BCB tuning: Focus server-side date filters, faster SGS failures, SICOR $select paging
• Packaging: agrobr health / agrobr doctor now work from the PyPI wheel