{"version":"2.1.310","anchor":"agent-summary-label-language-follows-summarywriter-model","canonical_anchor":"agent-summary-label-language-follows-summarywriter-model","heading":"Agent summary labels pick their language differently with Haiku","tier":"notice","area":"Subagents","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.310\/e\/agent-summary-label-language-follows-summarywriter-model","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.310","markdown":"### Agent summary labels pick their language differently with Haiku\n\nWhen Haiku writes an agent's summary label, it now uses the task's language instead of the main language of the notes\n\n**What**\n\nClaude Code writes a short label summarising what an agent is doing. When the model writing that summary is Haiku, the label now uses the language of the task. With other models it still uses the predominant language of the notes.\n\n**Why**\n\nThe language a summary label appears in can differ from before when Haiku writes it.\n\n- Area: Subagents\n- Tier: You'll notice\n- Useful: 1\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}