Validate precipitation probabilities and ice wording

docs/internal/facts.md

@@ -26,6 +26,8 @@ uses half-open period overlap to select hourly, narrative, daily, and alert
 data; it also derives precipitation timing. Convective outlooks are retained
 only when their valid interval overlaps the report period, with discussions
 kept for represented outlook days. Both collections are sorted deterministically.
+It rejects collected hourly data with a precipitation probability outside the
+finite 0 through 100 percentage domain before constructing derived facts.
 
 Report identity controls the summary shape:
 

docs/internal/forecast-derivation.md

@@ -26,12 +26,18 @@ its selected hourly periods and derives temperature and apparent-temperature
 ranges, timed precipitation and wind maxima, dominant and notable conditions,
 and weather indicators.
 
+Hourly precipitation probabilities must be finite percentages from 0 through
+100. Daily-summary construction rejects invalid values before they can affect
+timed maxima or precipitation windows.
+
 Indicators are deterministic checks over normalized values and condition text:
 heat, cold, and wind use package-owned numeric cutoffs; snow, ice, fog, and
 wind text are detected from the forecast description. `BuildPrecipTiming`
 sorts periods, records the maximum and first precipitation, groups contiguous
 periods at or above its package-owned probability threshold, and records
 thunder mentions.
+Ice detection uses the bounded condition vocabulary `ice`, `icy`, `freezing`,
+and `sleet` as whole words.
 
 Daypart temperature and apparent-temperature ranges are Fahrenheit values, and
 timed wind maxima are mph values. When only metric source fields are present,