Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.gdeltcloud.com/llms.txt

Use this file to discover all available pages before exploring further.

Overview

The GDELT Cloud MCP server exposes non-GDELT Cloud categories as enrichment. Use them after GDELT Cloud Events, Stories, or Entities establish the core signal.
CategoryWrapper ToolsUse
Macro financemacro_finance_tool_list, macro_finance_tool_get, macro_finance_tool_callQuotes, historical series, FX, commodities, rates, economic indicators, and symbol/search/catalog helpers.
Prediction marketsprediction_market_tool_list, prediction_market_tool_get, prediction_market_tool_callScenario markets, event pages, contract records, probabilities, status, and public market URLs when available.
Web researchweb_research_tool_list, web_research_tool_get, web_research_tool_callSearch and extraction for official statements, specialist context, corroboration, and source-article reading.

Shared Wrapper Parameters

WrapperParameterRequiredDescription
*_tool_listnonenoLists available nested tools in that category.
*_tool_gettool_nameconditionalSingle nested tool name to inspect.
*_tool_gettool_namesconditionalList of nested tool names to inspect together.
*_tool_calltool_nameyesExact nested tool name to execute.
*_tool_calltool_argumentsnoObject containing the selected nested tool’s parameters.

Macro Finance

Macro-finance nested tools are discovered dynamically through macro_finance_tool_list and macro_finance_tool_get. Use for:
  • equity and index quotes
  • historical price series
  • FX pairs
  • commodities
  • rates
  • economic indicators
  • provider symbol/search/catalog helpers when exposed
Provider symbols are not interchangeable across data vendors. Inspect the target schema first, then resolve provider-supported symbols through a search/catalog helper or web research. If a provider error returns retry guidance, change the lookup instead of retrying unchanged. When a selected macro-finance tool returns a recognizable historical payload, the wrapper normalizes it into top-level chart fields:
  • symbol / instrument
  • latest_value and latest_date
  • series: recent { label, date, value } points, usually daily observations
  • series_observation_count
Brief agents should prefer real provider series from these fields over synthetic two-point lines. Current quote fields are useful context, but they are not factual confirmation that a geopolitical or operational event caused the market move.

Prediction Markets

SEARCH_RELEVANT_MARKETS

Use when starting from a plain-language scenario and looking for likely matching contracts.
ParameterRequiredDescription
queryyesSignal, scenario, or event description to match against prediction markets.
categorynoOptional category filter.
statusnoOptional status filter. Default open.
limitnoCandidate markets to return, 1-25. Default 10.
include_historynoAttach recent daily probability history to top returned contracts when available. Default true.
history_daysnoProbability-history window, 2-30 days. Default 14.

SEARCH_EVENTS

Use for higher-level event pages that group related contracts.
ParameterRequiredDescription
searchnoFree-text search string.
categorynoOptional event category filter.
statusnoOptional status filter.
series_tickernoOptional series ticker filter.
limitnoMaximum events, 1-100. Default 10.
with_nested_marketsnoInclude nested markets. Default true.

SEARCH_MARKETS

Use for direct contract search by text or known identifiers.
ParameterRequiredDescription
searchnoFree-text search string.
statusnoOptional status filter.
event_tickernoOptional event ticker filter.
series_tickernoOptional series ticker filter.
limitnoMaximum markets, 1-100. Default 10.
include_historynoAttach recent daily probability history to top returned contracts when available. Default true.
history_daysnoProbability-history window, 2-30 days. Default 14.

GET_MARKET

ParameterRequiredDescription
tickeryesPrediction-market ticker returned by a search tool.
include_historynoAttach recent daily probability history when available. Default true.
history_daysnoProbability-history window, 2-30 days. Default 14.
Prediction-market records may include series, probability_series, and series_observation_count. Treat these as market-implied scenario context only. They can support a probability-context plot, but they never confirm that an event occurred.

Web Research

SEARCH_WEB

ParameterRequiredDescription
queryyesWeb search query.
search_depthnobasic or advanced. Defaults to basic.
max_resultsnoSearch results to return, 1-20. Wrapper normalizes provider fallback to a smaller safe value when needed.
time_rangenoRecency window such as day, week, month, year, 7d, or 30d.

EXTRACT_WEB_PAGES

ParameterRequiredDescription
urlsyesOne or more URLs to extract.
querynoOptional reranking query.
extract_depthnoExtraction depth. Default basic.
formatnoOutput format. Default markdown.
include_imagesnoInclude image URLs. Default false.
include_faviconnoInclude favicon URLs. Default false.

Use Pattern

  1. Anchor the analysis in GDELT Cloud structured Events, Stories, and Entities.
  2. Use macro finance when there are price, FX, rates, commodity, or economic-indicator implications.
  3. Use prediction markets when the question involves odds, policy, elections, escalation, or whether a tradable market exists.
  4. Use web research for official language, corroboration, specialist sources, and source-article extraction.