Open-Data MCP
Unify akshare, yfinance, EDGAR and more into one semantic schema so AI agents query financial and economic data by concept and entity — from a hundred sources, through one interface.
▶ Try it livefd-open-data-mcp is the core of the Wire line (寻数·柏讯) — let AI read the world for you. It is an open-data ontology MCP that unifies multiple Python data libraries into a single semantic schema, so AI agents query financial and economic data by concept and entity instead of adapting to each source’s API. A crawl platform runs behind it, feeding the ontology from a hundred sources.
One semantic schema, many sources
akshare, yfinance, EDGAR, World Bank, CNStats and other Python data libraries
are mapped into one semantic schema — semantic_observations. Agents ask in
terms of concepts and entities: Moutai’s price.close, China’s GDP. The
system resolves each query across sources, ranks them by quality, fails over
when a source is unavailable, and refreshes on a schedule.
An open-source fd-* family
The fd-* package family is open source: installable from PyPI, with the code on the FindDataTechnology GitHub org.