Block a user
correlator+summarizer: per-scene call-doc storage, fixes #96 and #114's real fix
Scene over-splitting: one radio conversation becomes two incidents
Closing per the investigation above: #5's stated mechanism (one call, transcript splits into 2+ scenes, each opens an incident) does not reproduce in 3 live dumps (0/10 candidate hits show…
Scene over-splitting: one radio conversation becomes two incidents
Same call_id linked into 2 incident records with identical extracted data (~2% of linked calls)
Scene over-splitting: one radio conversation becomes two incidents
Measured against 3 real dumps (corr_dump_9-7_0437am, corr_dump_9-7_pm, corr_dump_9-12) — the exact mechanism #5 describes does not reproduce
Scripted a scan for the literal pattern: one…
intelligence: shadow-mode upstream dispatch-vs-chatter classifier (#127)
Upstream dispatch-vs-chatter classifier — shadow mode
logan
deleted branch feat/115-chatter-classifier-shadow-mode from logan/server-26
2026-09-13 12:43:19 -04:00
Upstream dispatch-vs-chatter classifier — shadow mode
Review found and fixed two real gaps in the shadow-mode PR (#128, commit 3ae0bb2):
- Placement bug (would have made the shadow data itself misleading): the classifier ran after the…
logan
pushed to feat/115-chatter-classifier-shadow-mode at logan/server-26
2026-09-13 12:42:48 -04:00
ci: prune docker images before AND after deploy, not only on pull failure (#129)
Deploy failed with VM disk full — git pull could not write objects; prune ran too late to prevent it
ci: prune docker images before AND after deploy, not only on pull failure (#129)