{"version":"2.1.260","anchor":"narration-generation-now-distinguishes-a-truncated-outcome","canonical_anchor":"narration-generation-now-distinguishes-a-truncated-outcome","heading":"Narration generation now distinguishes a 'truncated' outcome","tier":"internal","area":"Background Tasks","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.260\/e\/narration-generation-now-distinguishes-a-truncated-outcome","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.260","markdown":"### Narration generation now distinguishes a 'truncated' outcome\n\nStatus-line narration that gets cut off by max tokens is now tracked as its own outcome.\n\nWhen the model call that generates the status-line narration hits `stop_reason: \"max_tokens\"`, it's now classified as a distinct `truncated` outcome with its own telemetry, rather than being treated as a generic failure.\n\n- Area: Background Tasks\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 1\/5"}