mcpField
G
GTFS Scorecard

GTFS Scorecard

Local
gtfsscorecard.orgMaps & Travel1Updated today

About GTFS Scorecard

GTFS Scorecard scores public GTFS Schedule feeds and translates validator findings into plain-language fixes. Transit agencies and feed reviewers can use its CLI to score a feed and its read-only MCP server to inspect published grades, fixes, and history.

Key features

  1. 01Scores GTFS Schedule ZIP feeds
  2. 02Prints grades and category scores
  3. 03Translates validator findings into fixes
  4. 04Searches tracked agencies and scorecards
  5. 05Reads grade history

Use cases

  1. 01Checking a GTFS feed before publication
  2. 02Finding plain-language fixes for validator notices
  3. 03Reviewing an agency's scorecard and grade history

MCP support

The read-only stdio MCP server lets clients search tracked agencies, retrieve scorecards and fixes, read grade history, and explain validator findings. It reads published scorecards and requires no key.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
"mcpServers": {
"chelseakr-gtfs-scorecard": {
"command": "uvx",
"args": [
"gtfs-scorecard"
]
}
}
}
Run commanduvx gtfs-scorecard

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
pypigtfs-scorecard1.5.1

Tools

9

Saved tool list · updated 2w ago

ParameterTypeRequiredDescription
gradestringNoLetter grade A-F
limitintegerNoMax results (default 20)
querystringNoName fragment or exact agency id
statestringNoLegacy state/province-name filter; prefer subdivision
countrystringNoISO 3166-1 alpha-2 country code, e.g. CA
subdivisionstringNoISO 3166-2 code or subdivision name, e.g. CA-ON or Ontario

Details

From the project README

gtfs-scorecard

The scoring pipeline, command-line tool, and read-only MCP server behind GTFS Scorecard, which grades public GTFS Schedule feeds and turns validator notices into plain-language fixes.

What it does

  • scorecard try <feed-url> scores one GTFS Schedule zip with the MobilityData validator (Java 17 or newer must be on PATH) and prints the grade, the category scores, and the top fixes.
  • scorecard-mcp is a read-only Model Context Protocol server over stdio. Its tools read the published scorecards on gtfsscorecard.org: search tracked agencies, get a scorecard and its fixes, read a feed's grade history, and explain a validator finding. It needs no key and has no write surface.
  • gtfs-scorecard is the same MCP server under the distribution's name, so uvx gtfs-scorecard starts it.

The Python import package is scorecard_pipeline. The distribution was named scorecard-pipeline until 1.5.1; PyPI refused that name as too similar to an existing project, and nothing was uploaded under it.

Install

uvx gtfs-scorecard
uvx --from gtfs-scorecard scorecard try https://example.org/gtfs.zip
pip install gtfs-scorecard

The first starts the MCP server. The second runs the command-line tool, which has its own name, so uvx needs --from to find the package that provides it. Check https://pypi.org/project/gtfs-scorecard/ for the versions that exist. To run code that is not released yet, install from the repository:

uvx --from "git+https://github.com/ChelseaKR/gtfs-scorecard#subdirectory=pipeline" scorecard-mcp

Connect an assistant

An MCP client config that starts the server from PyPI:

{
  "mcpServers": {
    "gtfs-scorecard": {
      "command": "uvx",
      "args": ["gtfs-scorecard"]
    }
  }
}

Set SCORECARD_BASE_URL to point it at a fork or a local preview.

More

Apache-2.0.

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.