{"version":"2.1.314","anchor":"auto-mode-classifier-context-gains-text-provenance-and-a-han","canonical_anchor":"auto-mode-classifier-context-gains-text-provenance-and-a-han","heading":"Auto mode's safety check receives more detail, and its context file is renamed in a dialog","tier":"notice","area":"Auto Mode","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.314\/e\/auto-mode-classifier-context-gains-text-provenance-and-a-han","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.314","markdown":"### Auto mode's safety check receives more detail, and its context file is renamed in a dialog\n\nThe auto mode safety check now receives information about where text came from, and a dialog calls its document the 'auto mode context'\n\n**What**\n\nIn auto mode, Claude Code runs a check (a classifier) that decides whether an action can go ahead without asking you. The information sent to that check now includes:\n\n- details about where pieces of text came from\n\n- a flag saying whether structured output is being handed back\n\nThe step that takes over pending actions now also covers background tasks.\n\nThe confirmation dialog for replacing the built-in environment description now calls that document the \"auto mode context\", the text that defines trusted hosts, sensitive targets and repository scope. It used to call it the \"classifier context\".\n\n**Why**\n\nThis changes what the auto mode check sees when it makes decisions. The dialog's new name matches the term \"auto mode\" that you already see elsewhere.\n\n- Area: Auto Mode\n- Tier: You'll notice\n- Useful: 1\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}