# who-gho-mcp-server [Health: Active]

**Category:** 🧬 Biology & Bioinformatics  
**Repository:** https://github.com/cyanheads/who-gho-mcp-server  
**GitHub Stars:** 3  
**npm Downloads (last month):** 1195  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/who-gho-mcp-server

## Description
WHO Global Health Observatory — 3,059 indicators across 194 member states.

## Tools
Capabilities this server exposes over MCP:

- **who_search_indicators** — Search the GHO indicator catalog by keyword in indicator names
- **who_list_indicators** — Browse the full indicator catalog with pagination
- **who_get_indicator_metadata** — Fetch indicator names and dimension types, processing 10 codes per call with continuation for larger batches
- **who_list_dimensions** — List all dimension type codes available in the GHO API
- **who_list_dimension_values** — List valid codes and labels for a dimension type (COUNTRY, REGION, SEX, etc.)
- **who_query_indicator_data** — Query data rows for an indicator with spatial, temporal, and dimension filters

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

```json
"mcpServers": {
  "who-gho-mcp-server": {
    "command": "bunx",
    "args": ["@cyanheads/who-gho-mcp-server@latest"]
  }
}
```

## Documentation

## What who-gho-mcp-server does

who-gho-mcp-server exposes the WHO Global Health Observatory (GHO) through the Model Context Protocol. It covers indicators from WHO member states and supports discovery before querying: an agent can search indicator names, browse the catalog, inspect supported dimensions, enumerate valid dimension values, and retrieve matching data rows.

The server is suited to questions that require structured health statistics by geography, year, sex, age group, or income grouping. It distinguishes country rows from WHO regional aggregates and World Bank income-group aggregates, allowing queries to target the appropriate spatial level.

## How it works

Indicator discovery starts with `who_search_indicators` or `who_list_indicators`, which return indicator codes and names for later use. `who_get_indicator_metadata` reports an indicator's name and available dimension types. Metadata may contain no listed dimensions for some indicators; in that case, the response can include a note based on a sample data row.

Dimension discovery uses `who_list_dimensions` followed by `who_list_dimension_values`. Dimension values include codes and labels, and hierarchical results can include parent information. For example, a country lookup can be narrowed by a WHO region code.

`who_query_indicator_data` accepts one spatial filter category per call: country codes, region codes, or income-group codes. It also supports year ranges, positional dimension values, optional sex selection, uncertainty bounds, sorting, and offset pagination. Returned dimension type/value pairs help identify positional filters for subsequent queries.

## Setup and configuration

The package is available as `@cyanheads/who-gho-mcp-server` and can be installed through the MCP setup flows offered for Claude Desktop, Cursor, and VS Code. It supports both STDIO and Streamable HTTP. A hosted endpoint is available at `https://who-gho.caseyjhand.com/mcp`, while local deployments can use the npm package or the published Docker image.

The server also ships an MCPB bundle and an agent skill named `who-gho-guide`. The skill covers dimension-position discovery, complete country comparisons, and choosing between country and regional aggregate rows. MCP clients can discover the skill through the server's skill and resource interfaces.

## Tools and capabilities

- Search indicator names with word-based matching and paginated results.
- Browse the complete indicator catalog with offset pagination.
- Resolve indicator metadata in batches, with continuation for unprocessed codes.
- List GHO dimension types and their valid codes and labels.
- Filter hierarchical dimension values by a parent code.
- Query indicator rows with spatial, temporal, and dimension filters.
- Read indicator metadata and dimension values through `who://` resources.
- Request uncertainty intervals, ascending or descending year order, and paged result sets.

## Limitations and notes

Search matches whitespace-separated words as name substrings; it does not perform stemming or fuzzy expansion. Indicator metadata does not reveal dimension positions, so query positions should be inferred from returned type/value pairs or the accompanying guidance.

Spatial filter categories are mutually exclusive. Queries can fail for unknown indicators, unavailable data, ambiguous spatial filters, invalid year ranges, malformed identifiers, or other invalid parameters. Large catalog, metadata, dimension, and data results require pagination. Metadata accepts up to 1,000 codes, but processing is batched and may return codes for a later continuation call.

_Full upstream README: https://allmcps.com/mcp/who-gho-mcp-server/readme_