Platform
·BuiltDendrite
The corpus engine behind the fleet, one config turns a raw dataset into an MCP server, a REST API, and a static site.
Focus
Shared framework
Dendrite is the template every government-data product is built on. Describe a dataset once, its schema, sources, and the questions people ask of it, and Dendrite generates a typed ingestion pipeline, an MCP server for agents, an OpenAPI REST layer for apps, and a set of GEO-optimized landing pages. It is the reason a new federal dataset becomes a production service in days instead of months.
Highlights
- ▹Single corpus.config.ts drives MCP tools, REST routes, schema, and pages
- ▹Shared ingestion, normalization, and entity-resolution layer reused across 40 products
- ▹443 MCP tools and 313 REST routes generated, not hand-written
- ▹Static GEO pages so each dataset is discoverable without an agent
Why a framework
The first two datasets were built by hand. By the third it was obvious the work was the same shape every time: pull a canonical source, normalize it, expose it three ways, and make it searchable. Dendrite captures that shape so the only real work left is understanding the data.