Data Nj

What the Data Nj MCP server does

The Data Nj MCP server connects MCP clients to the New Jersey Open Data catalog at data.nj.gov, which is powered by Socrata. It provides three focused tools for discovering datasets, understanding their structure, and retrieving records.

datasets searches the catalog by keyword and returns each matching dataset's resource ID, name, description, category, and update date. The resource ID is the key needed by the other tools. metadata accepts a resource ID and returns schema information, column types, row count, category, and last-updated details. query runs a Socrata SoQL request and returns matching rows as JSON.

This fits workflows that need public New Jersey data in a structured form, such as inspecting a dataset before analysis or asking an agent to retrieve records matching specific conditions.

How it works

A typical interaction starts with datasets, using a keyword such as crime to locate relevant catalog entries. After selecting a resource ID, call metadata to learn the available columns and their types. That information can guide a later query call.

Queries accept SoQL clauses through where, select, group, and order. These values are supplied without the leading dollar sign used in some Socrata URL examples. limit and offset support bounded results and pagination. The response contains the rows returned by the New Jersey Open Data service as JSON.

The hosted endpoint also exposes shared Pipeworx gateway tools in addition to the pack-specific tools. Its initialization response is the authoritative source for the tools available at connection time. The local package exposes only the three Data Nj tools.

Setup and configuration

For a hosted connection, add the Pipeworx MCP endpoint to a client configuration:

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

The package can also run locally over stdio with npx:

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

The local mode does not require a Pipeworx account or a gateway round trip. The hosted option can also be used through the full Pipeworx gateway. No environment variables are documented for either setup.

Tools and capabilities

• datasets: Find New Jersey Open Data catalog entries by keyword.
• metadata: Inspect a dataset's columns, types, row count, category, and update date.
• query: Execute SoQL selection, filtering, grouping, ordering, and pagination against a resource ID.
• Hosted access: Call the tools through Pipeworx's streamable MCP endpoint.
• Local access: Run the package as a stdio MCP process.

Limitations and notes

The tools require a Socrata resource ID for metadata and data queries; a catalog search is the normal way to find one. Query syntax follows Socrata SoQL, and the clause arguments omit the leading $.

The hosted pack endpoint includes Pipeworx shared meta-tools such as routing and discovery tools, so its tool list is larger than the three pack tools. The standalone process does not include those shared tools.

The Data Nj MCP server is focused on data.nj.gov. It should not be treated as a general-purpose database connector or as an interface for arbitrary Socrata domains unless those domains are explicitly supported separately.