Census Tigerweb MCP server

@pipeworx/census-tigerweb

The Census Bureau's TIGERweb ArcGIS REST service: every TIGER/Line geography the Bureau publishes as a queryable layer — census tracts, block groups, 2020 blocks, ZCTAs, congressional and state legislative districts, school districts, tribal areas, places, counties, urban areas and PUMAs, with their GEOIDs, names, land/water area and interior points.

Part of Pipeworx — an MCP gateway connecting AI agents to 1679+ live data sources.

Tools

queryable layers in a TIGERweb service. Call this first; layer ids move between vintages.

• tigerweblistlayers(service?, filter?, includelabellayers?) — the

ArcGIS WHERE-clause query against any layer.

• tigerwebquery(layerid, service?, where?, outfields?, limit?, offset?, orderby?, return_geometry?) —

geography containing a coordinate, in one call.

• tigerwebgeographyat_point(latitude, longitude, service?, layers?) — every

Auth

Keyless.

Data sources

• — current-vintage geography (default).
• — the service directory, including ACS and Census2020 vintages.

Companion to census-geocoder: the geocoder maps an address to a tract, this pack lists and describes the geographies themselves.

Traps

annotation, not boundaries. Querying one returns rows that look like real geography. tigerweblistlayers hides them unless asked.

• Half of the 80 layers in tigerWMS_Current are ... Labels layers — point

from another service.

• Layer numbers move between services and vintages. Never carry one over

does not. A wrong-typed WHERE clause returns zero rows with a 200.

• FIPS fields are zero-padded strings: STATE='11' matches, STATE=11

layers do not share a schema (Census Tracts has no BLOCK, ZCTAs have no COUNTY). tigerwebgeographyatpoint asks for and trims afterwards.

• ArcGIS errors on an outFields name a layer does not have, and these

more and you silently get the service max.

• resultRecordCount caps a page; the service has its own max too. Ask for

• Geometry is large — return_geometry defaults to false.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "census-tigerweb": {
      "url": "https://gateway.pipeworx.io/census-tigerweb/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/census-tigerweb/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — askpipeworx, discovertools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1679+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

Shortened. The full README is on GitHub.