A macro data layer that speaks agent.

MacroFeed turns fragmented release calendars and official observations into focused tools with predictable, structured responses.

Small tools. Clear intent.

Each public method is read-only and scoped to a recognizable agent task, reducing discovery noise and unnecessary recovery calls.

QUERY / MCP

get_upcoming_events

Return bounded scheduled or released events by date, currency, country, impact, or indicator.

RESOLVE / MCP

get_next_release

Find the next CPI, employment, GDP, or central-bank release with one focused call.

INSPECT / MCP

get_event

Retrieve a complete normalized event by its stable event identifier.

RESEARCH / MCP

get_release_history

Read historical actual and previous observations without inventing unverified event timestamps.

VERIFY / MCP

get_coverage_summary

Inspect event counts, date ranges, currency coverage, and field-density metrics.

DISCOVER / MCP

list_calendar_sources

Understand the official institutions and source methods behind the feed.

One event, fully traceable.

The normalized record keeps market context, timing, availability, source identity, and retrieval metadata together.

SCHEDULEDOFFICIAL CALENDAR
USD

Consumer Price Index

HIGH IMPACT
JUL 1408:30ET / 12:30 UTC
Actual
Forecast
Previous
Period
2026-06
SOURCEU.S. BUREAU OF LABOR STATISTICS

More intelligence, without blurred labels.

Revision history and standardized surprise analytics are part of the M2 direction. Consensus, model estimates, official projections, and market-implied data will remain separate types.

M2 / IN PROGRESS

Revision history

Preserve the path from first release to the latest official value.

M2 / IN PROGRESS

Surprise analytics

Compare actual versus attributed consensus and normalize errors by indicator.

POST-M2 / PLANNED

Macro context

Summarize nearby events without making unsupported causality claims.