Tool ReferenceOther
data_dictionary
Data Dictionary
What weather data exists for a topic, and which tool serves it. Free, no credits.
Call this FIRST for exploratory questions ("do you have hail data?",
"how far back does rainfall go?"). Three modes:
1. SEARCH (query given — the primary mode): searches every variable/product
across all datasets (station, gridded 15km, climate normals, precipitation,
severe weather, hurricanes, risk scores, forecast, ...). Returns compact
routing entries {variable, dataset, units, tool, one_line} — call the
named tool next. detail='full' adds resolution, coverage, and caveats.
2. LOOKUP (variable_names given): definition, units, and default flag for
exact API variable names (e.g. 'surfaceTemperature'). Not a search —
names must match exactly.
3. LIST (neither given): dataset='station'|'gridded'|'climate_normals'
returns that dataset's field packages, valid fields, and supported
intervals — call before station_historical_data, station_climate_normals,
or gridded_history_15km to choose the 'fields' parameter (if the intent
doesn't map to a package, use fields='all' in the data call).
dataset='all' (default) returns the compact dataset catalog.https://mcp.barocline.com/mcphttps://mcpkey.barocline.com/weather-risk/mcphttps://mcpoauth.barocline.com/mcpRead-onlyIdempotent
Parameters
| Name | Type | Required | Default |
|---|---|---|---|
query | string | null | No | null |
dataset | string | No | "all" |
variable_names | string | null | No | null |
detail | string | No | "compact" |
top_k | integer | No | 5 |
content_mode | string | No | "text" |
Sample input
{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "data_dictionary",
"arguments": {
"dataset": "all",
"detail": "compact",
"top_k": 5,
"content_mode": "text"
}
}
}{
"name": "data_dictionary",
"arguments": {
"dataset": "all",
"detail": "compact",
"top_k": 5,
"content_mode": "text"
}
}