mcpField
T
Tx Dmv

Tx Dmv

Remote
pipeworx.ioLegal & Government0Updated 5w ago

About Tx Dmv

Tx Dmv is an MCP server for querying Texas vehicle registration totals by fiscal year. It reports statewide counts for cars, light pickups, and motorcycles, with coverage limited to FY2001–FY2021.

Key features

  1. 01Fiscal-year vehicle totals
  2. 02Passenger car, pickup, and motorcycle counts
  3. 03Fleet-share percentages
  4. 04Coverage and refresh-date details

Use cases

  1. 01Compare Texas vehicle totals across fiscal years
  2. 02Check statewide motorcycle or pickup registrations

MCP support

Provides the `tx_dmv_vehicle_registrations` tool to MCP clients through Pipeworx or as a standalone stdio server. It can also be called over HTTP.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://gateway.pipeworx.io/tx-dmv/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

@pipeworx/tx-dmv

Texas DMV (TxDMV) MCP — statewide vehicle, pickup and motorcycle registration totals by fiscal year, from TxDMV's own published series.

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

Tools

  • tx_dmv_vehicle_registrations(fiscal_year?, limit?) — total vehicles registered in Texas for a fiscal year, split into passenger cars, pickup trucks of one ton or less and motorcycles, each with its share of the fleet. Answers "how many vehicles are registered in Texas", "how many motorcycles are registered in Texas", and growth across years.

Auth

Keyless.

Grain and vintage — read this before quoting a number

Two things about this dataset trip people up, so both are reported on every response:

  1. Statewide only. TxDMV publishes one row per fiscal year for the whole state. There is no county, city or ZIP breakdown in this series, and no make/model/fuel split. Passing county, zip, make, model_year or fuel_type returns { found: false, reason: "grain_unavailable" } with a hint rather than silently ignoring the argument. For a ZIP-level fleet breakdown, ca_dmv_vehicle_registrations covers California at ZIP × make × model-year × fuel.
  2. The series stops at FY2021. TxDMV last refreshed it on 2023-03-06, so it covers fiscal years 2001 through 2021 and recent years are simply absent. as_of carries the refresh date, covers_fiscal_years the span, and a note says so in words. Asking for fiscal_year="2024" returns fiscal_year_unavailable rather than an empty row list.

total_vehicles is only set when a single fiscal year is requested, because that row genuinely is the whole Texas fleet for that year. Rows from different fiscal years are not summed: they are cumulative snapshots of largely the same vehicles, so a sum would double-count every car that stayed registered.

Verified live 2026-07-29: FY2001 total_all_vehicles_registered = 17,906,116; FY2021 = 25,236,442.

Data sources

  • https://data.texas.gov/resource/cbmj-zeje.json — TxDMV "Passenger Vehicle, Motorcycle, and Pickup Truck Registrations by Year". 21 rows, FY2001–FY2021.
  • https://data.texas.gov/api/views/cbmj-zeje.json — rowsUpdatedAt, used for as_of.

The three *_of_all_vehicles columns are fractions of 1 (0.5831), not percentages; the pack converts them to percentages. fiscal_year is stored as text, so it is filtered with a string equality and ordered as a string — which is safe for four-digit years.

Quick Start

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

{
  "mcpServers": {
    "tx-dmv": {
      "url": "https://gateway.pipeworx.io/tx-dmv/mcp"
    }
  }
}
What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/tx-dmv/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, 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

curl -X POST https://gateway.pipeworx.io/v1/tools/tx_dmv_vehicle_registrations \
  -H 'Content-Type: application/json' \
  -d '{"fiscal_year":"2021"}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/tx_dmv_vehicle_registrations. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Standalone (no gateway account)

This package also runs as a local stdio MCP server — no Pipeworx account, no gateway round-trip:

{
  "mcpServers": {
    "tx-dmv": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-tx-dmv"]
    }
  }
}

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-tx-dmv

It speaks MCP over stdin/stdout and answers initialize/tools/list/tools/call for only this pack's tools — none of the shared meta-tools the gateway connection above adds. Same source, same tools, no ask_pipeworx routing.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Tx Dmv data" })

The gateway picks the right tool and fills the arguments automatically.

More

License

MIT

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.