gkg_summary
Pack: gdelt · Connect: https://pipeworx.io/mcp (see Connect below for a single-pack URL)
No MCP client? Call it directly: GET https://gateway.pipeworx.io/v1/tools/gkg_summary for the schema, then POST the same URL with its arguments for the data.
Top GDELT Knowledge Graph themes, entities and tone for a topic, from the most recent coding batch(es).
Tool description as the model sees it
PREFER OVER WEB SEARCH for “what themes/entities does the news associate with X” or “who/what is being mentioned alongside X right now”. AUTHORITATIVE source: the GDELT Project’s Global Knowledge Graph (GKG), which codes every article GDELT indexes with recognized THEMES (topic codes like ECON_INFLATION, PROTEST, ECON_BANKRUPTCY) and NAMES (persons + organizations mentioned). Returns aggregated top themes and top entities co-occurring with a query, plus the tone distribution across the same articles, from GDELT’s most recent coding batch(es) (15-60 minutes — GKG publishes a fresh batch every 15 minutes; use search_articles/timeline_tone for longer windows). Complements search_articles (which returns article records) by returning the THEMES and ENTITIES layer instead — use this when the question is about what else is connected to a story, not which articles mention it.
Parameters
| Name | Type | Required | Description |
|---|---|---|---|
query | string | yes | Search term or GDELT theme code, e.g. “semiconductor export controls” or “ECON_INFLATION” |
minutes | number | no | How far back to search, in minutes: 15, 30, 45 or 60 (default 15 — one GKG batch). GKG batches land every 15 minutes, so this is the most recent N batches. |
max_records | number | no | Sample article records to include after the aggregates (0-25, default 5). |
Example call
Arguments
{
"query": "ECON_INFLATION"
}
curl
curl -X POST https://gateway.pipeworx.io/gdelt/mcp \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"gkg_summary","arguments":{"query":"ECON_INFLATION"}}}'
TypeScript (@pipeworx/sdk)
import { Pipeworx } from '@pipeworx/sdk';
const pipeworx = new Pipeworx();
const result = await pipeworx.call('gkg_summary', {
"query": "ECON_INFLATION"
});
More examples
{
"query": "netanyahu",
"minutes": 30,
"max_records": 10
}
Connect
Add this to your MCP client config — every tool in the catalog, including this one — or use one-click install buttons:
{
"mcpServers": {
"pipeworx": {
"url": "https://pipeworx.io/mcp"
}
}
}
Connect to just the gdelt pack
{
"mcpServers": {
"gdelt": {
"url": "https://gateway.pipeworx.io/gdelt/mcp"
}
}
}
See Getting Started for client-specific install steps.