{"version":"2.1.315","anchor":"agent-summary-label-always-uses-the-task-language","canonical_anchor":"agent-summary-label-always-uses-the-task-language","heading":"Background agent summary labels are always written in the task's language","tier":"notice","area":"Subagents","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.315\/e\/agent-summary-label-always-uses-the-task-language","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.315","markdown":"### Background agent summary labels are always written in the task's language\n\nThe short label summarising an agent's work is now always written in the language of the task, for every model\n\n**What**\n\nClaude Code writes a short label summarising what an agent (a helper that works on part of a task) is doing. That label is now always written in the language of the task. Previously, for models other than Haiku, it was written in the main language of the agent's notes.\n\n**Why**\n\nThe language of these labels is now more predictable.\n\n- Area: Subagents\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}