GEFRI MCP server

GEFRI provides a public, read-only Model Context Protocol endpoint for AI assistants at https://gefri.educationfutures.com/api/mcp. It uses Streamable HTTP, requires no authentication, and exposes only published GEFRI data and documentation.

Historical queries use annual December snapshots beginning in 2016. The current snapshot keeps its actual build month. Confidence labels describe the completeness and directness of the evidence; indicator results separately identify original and imputed observations.

Available tools

get_countryCountry scores, ranks, evidence confidence, and indicator provenance.
list_countriesSearch and filter countries by region, income, FCS status, score, or confidence.
get_rankingsGlobal or regional rankings for the composite or any dimension.
compare_countriesSide-by-side comparison of two to ten countries.
get_country_trendAnnual December history from 2016 plus the exact current build period.
analyze_regionRegional summaries, with unweighted averages by default and optional population weighting.
analyze_cohortAnalysis of income, FCS, regional, or custom country cohorts.
get_indicatorDefinitions, source attribution, transformation, coverage, and imputation details.

Using regional analysis

Call analyze_region with a GEFRI region name or a familiar alias such as LAC, MENA, EAP, or SSA. Use weighting: population only when the question calls for a population-weighted result; otherwise retain the default unweighted country average.

Responsible use and attribution

GEFRI is a comparative readiness index, not a direct measure of education quality or a prediction of future success. Cite Education Futures and the originating indicator agencies returned by the service. Public GEFRI data and documentation are available under CC BY 4.0.

For conventional JSON access, see the REST API documentation. For methodology and limitations, see the methodology and technical appendix.