Follow-up from #3791 plan (deferred scope item 6b).
#3791 makes the monitor-tick writer promote embedded measurements out of warnings tokens into typed sibling keys (e.g. low-disk 90% → warnings:["low-disk"], disk_free_pct:90) and normalizes warnings/actions to a closed vocabulary.
Once the writer emits the canonical schema, .claude/skills/daily-summary/SKILL.md should be updated to aggregate on the closed vocabulary + typed measurement fields (instead of hand-clustering free-text spellings as it does today per amendment 2).
Scope
Why separate from #3791
Read-side change; #3791 is write-side only. Best sequenced after the writer canonicalization (this issue) and the corpus backfill are in place.
Parent: #3791.
Follow-up from #3791 plan (deferred scope item 6b).
#3791 makes the monitor-tick writer promote embedded measurements out of
warningstokens into typed sibling keys (e.g.low-disk 90%→warnings:["low-disk"], disk_free_pct:90) and normalizeswarnings/actionsto a closed vocabulary.Once the writer emits the canonical schema,
.claude/skills/daily-summary/SKILL.mdshould be updated to aggregate on the closed vocabulary + typed measurement fields (instead of hand-clustering free-text spellings as it does today per amendment 2).Scope
### Watch items,### Tick aggregates) to read the typed measurement keys and count warnings by canonical vocab token.Why separate from #3791
Read-side change; #3791 is write-side only. Best sequenced after the writer canonicalization (this issue) and the corpus backfill are in place.
Parent: #3791.