DataSF

Provides access to San Francisco open data via MCP with search, schema retrieval, and SoQL queries.

typescript

0

GitHub Stars

typescript

Language

5 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "fwextensions-datasf-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "github:fwextensions/datasf-mcp"
      ],
      "env": {
        "SOCRATA_APP_TOKEN": "your-optional-token"
      }
    }
  }
}

You can use the DataSF MCP Server to give AI assistants streamlined access to San Francisco’s open datasets. It hides the complexity of the Socrata API, provides smart schema handling, and lets you search, explore, and query datasets with SoQL through a consistent MCP interface.

How to use

Connect to the DataSF MCP Server from your MCP client using the stdio-based command configuration. You can run the server via npx and point your client at the latest build. Once connected, you can search datasets, retrieve schemas, and run SoQL queries against any dataset. Use the search endpoint to discover datasets by keywords, get_schema to inspect columns and types, and query_datasf to execute SoQL queries with optional automatic column name correction.

How to install

Prerequisites: you need Node.js 18 or higher and npm or yarn installed on your system.

# Install Node.js 18+ from the official source if needed
# Then install dependencies and prepare the MCP server

# Quick start: use the MCP configuration that pulls the server from GitHub via npx

Additional configuration and notes

The recommended way to run the server from a client is via npx, which downloads and runs the latest MCP server build directly. You can optionally provide a Socrata App Token to increase rate limits when accessing datasets that require authentication.

Environment variable you can set for higher rate limits:

SOCRATA_APP_TOKEN: your-optional-token

Example MCP configuration (stdio) to run the server with npx and the GitHub-hosted MCP build, including the optional token:

{
  "mcpServers": {
    "datasf": {
      "command": "npx",
      "args": ["-y", "github:fwextensions/datasf-mcp"],
      "env": {
        "SOCRATA_APP_TOKEN": "your-optional-token"
      }
    }
  }
}

Available tools

search_datasf

Search for datasets by keywords and return relevant results.

list_datasf

Browse datasets, optionally filtered by category.

get_schema

Fetch the schema for a specific dataset, including column names and data types.

query_datasf

Execute a SoQL query against a dataset with optional automatic column name correction.