{"version":"2.1.310","anchor":"memory-context-per-turn-wait-no-longer-reports-early-get-t","canonical_anchor":"memory-context-per-turn-wait-no-longer-reports-early-get-t","heading":"Memory context fetches at turn end are dropped when a new turn starts","tier":"internal","area":"Elsewhere","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.310\/e\/memory-context-per-turn-wait-no-longer-reports-early-get-t","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.310","markdown":"### Memory context fetches at turn end are dropped when a new turn starts\n\nWhen a new turn starts, Claude Code now abandons a memory context fetch still running from the last turn, and stops reporting early fetch timing\n\n**What**\n\nA turn is one exchange: you send a message and Claude answers. Claude Code fetches memory context, the remembered information it gives Claude, around each turn. If a fetch started at the end of the previous turn is still running when the next turn starts, Claude Code now drops it and clears its record of the fetch. It no longer keeps track of fetches started early.\n\nThe usage report Claude Code sends about waiting for memory context also no longer includes when the early fetch started or how long before the turn it began.\n\n**Why**\n\nThe timing of memory context fetches has been reworked, so an unfinished fetch from the last turn no longer carries over.\n\n- Area: Elsewhere\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}