Why this page exists
Time-sensitive workflows fail when code assumes data that doesn’t exist yet, or requests history that doesn’t go back far enough. This page gives the refresh cadence, history depth, and coverage for each asset so agents set correct temporal parameters.Freshness at a glance
Platform context: 1M+ data points refreshed monthly, 30 years of historical
trends, and 22,000+ verified property and public-record sources. Per-asset rows above
take precedence for any specific call.
Rules for time-aware tool calls
1
Match granularity to cadence
Monthly assets (Market, HPI) take month-level
YearMonth ranges. Never request daily or
real-time values — the data does not exist at that granularity.2
Omit time_range for latest data
For analytics tools, omit
time_range entirely to retrieve the latest available period.
Do not pass a partial object — both from and to are required when time_range is present.3
Distinguish actual vs. forecast
HPI forecasts and 30-year climate projections are model projections, not observations.
Always label them as forecasts or estimates when presenting to users.
4
Check the envelope messages
Freshness notes and partial-coverage warnings surface in the
messages array of the
response envelope. Always read these.