# cdc-health-mcp-server [Health: Active]

**Category:** 🧬 Biology & Bioinformatics  
**Repository:** https://github.com/cyanheads/cdc-health-mcp-server  
**GitHub Stars:** 5  
**npm Downloads (last month):** 1295  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/cdc-health-mcp-server

## Description
MCP server for discovering and querying CDC public health datasets via the Socrata SODA API.

## Tools
Capabilities this server exposes over MCP:

- **cdc_discover_datasets** — Search the CDC dataset catalog by keyword, category, tag, or asset type
- **cdc_get_dataset_schema** — Fetch column schema, row count, and metadata for a dataset
- **cdc_list_catalog_vocabulary** — List the catalog's category and tag values with their entry counts
- **cdc_query_dataset** — Execute SoQL queries — filter, aggregate, sort, full-text search, select fields
- **cdc_query_wonder** — Query CDC WONDER for national mortality, population, and death rates across five databases

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "cdc-health-mcp-server": {
    "command": "npx",
    "args": ["-y","@cyanheads/cdc-health-mcp-server"]
  }
}
```

## Documentation

## What cdc-health-mcp-server does

cdc-health-mcp-server makes CDC public-health data available through the Model Context Protocol. It covers the CDC Open Data catalog and CDC WONDER, with particular use for finding datasets, understanding their columns, querying tabular data, and examining national mortality statistics.

The catalog tools can search by keyword, category, tag, or asset type; inspect a dataset's metadata and schema; and enumerate catalog categories and tags with entry counts. Queryable Socrata datasets can be filtered, grouped, sorted, aggregated, selected by field, or searched with full-text criteria. The WONDER tool handles national mortality, population, and death-rate queries across five CDC databases.

Two MCP resources provide either the 50 most-viewed catalog entries or metadata and initial columns for a specified dataset. An `analyze_health_trend` prompt offers a guided workflow for investigating a public-health question. The package also includes the `cdc-health-guide` skill, which provides guidance for comparing rates, interpreting race and Hispanic-origin dimensions, working with provisional data, and handling stratified Socrata tables.

## How it works

The server can run over STDIO or Streamable HTTP. A hosted endpoint is available at `https://cdc.caseyjhand.com/mcp`; local installations can use the npm package, Docker image, or MCP bundle.

Catalog discovery returns asset type, permalink, column information, descriptions, counts, and applied filters. Dataset schemas include field names, data types, descriptions, row counts, and selected publisher metadata. Queries return the effective SoQL clauses and paging details. Dataset results are limited to 1,000 rows per call, while WONDER requests can return up to 10,000 rows and reject groupings that could exceed that size.

WONDER queries are always national and can group by dimensions such as year, month, age group, sex, race, Hispanic origin, ICD chapter or subchapter, cause, and injury intent. Available databases cover underlying and multiple causes of death from 1999–2020, single-race datasets from 2018–2024, and provisional statistics beginning in 2018. The provisional database can include a partial latest year.

## Setup and configuration

Install the npm package through an MCP client using the package's `npx` installation flow, or deploy the published Docker image. The repository also distributes an MCP bundle for clients that support bundle-based installation. Direct installation links are provided for Claude Desktop, Cursor, and VS Code.

For remote use, configure an MCP client with the hosted Streamable HTTP endpoint. For local use, select the STDIO transport and launch the package through the client. The server is implemented in TypeScript and uses the MCP SDK; its published runtime materials also include the packaged skill.

## Tools and capabilities

- `cdc_discover_datasets`: search the catalog with keyword, category, tag, and asset-type filters.
- `cdc_get_dataset_schema`: retrieve columns, row count, descriptions, and dataset metadata.
- `cdc_list_catalog_vocabulary`: list categories and ranked tags with dataset counts.
- `cdc_query_dataset`: issue SoQL selections, filters, grouping, sorting, aggregation, having clauses, and full-text searches.
- `cdc_query_wonder`: query five CDC WONDER databases for national mortality and related rates.

Dataset IDs use Socrata's four-by-four format. Schema columns and query results support paging. Stable offset paging for dataset queries requires an order ending with a unique tie-breaker, such as `:id` for ungrouped rows. WONDER supports ICD-10 codes, ranges, and lists, with database-specific race group definitions.

## Limitations and notes

Catalog charts, maps, and other non-tabular assets are not directly queryable; responses identify the underlying dataset where applicable. Catalog search pagination has a 10,000-entry boundary, and dataset query responses do not include a total count. WONDER results are national rather than geographic subdivisions, and race series from bridged-race and single-race database families should not be compared directly. The latest provisional year may be incomplete, and some causes can be withheld.

_Full upstream README: https://allmcps.com/mcp/cdc-health-mcp-server/readme_