Sugra API is intelligence infrastructure for AI agents and decision platforms: one key and one normalized format for cross-domain primary-source data, surfaced to agents through the Model Context Protocol. The connector exposes a small, stable tool surface over the full Sugra catalog of more than 1,300 endpoints aggregating 130+ primary sources across data domains including markets, macroeconomics and central banks, corporate filings, public finance, trade flows, agriculture, climate and natural events, maritime activity, prediction markets, global news, network intelligence, and entity screening. Main workflows: discover endpoints by natural-language search and inspect their schemas, run a one-step find-and-call to retrieve data, pull macroeconomic series and central-bank rates, fetch market prices, history and fundamentals through Sugra Finance, screen a person or organization name against sanctions corpora, resolve free text to a canonical entity, and compose a normalized current view or bounded timeseries for an entity with freshness and provenance. Works with Anthropic, OpenAI, Google, and xAI model families and any MCP-enabled client. A free Sugra account is the only connection requirement; the read-only surface never writes, deletes, sends, or publishes.