{"version":"2.1.299","anchor":"subagenthandoff-classifier-now-gets-inherited-verified-huma","canonical_anchor":"subagenthandoff-classifier-now-gets-inherited-verified-huma","heading":"Auto mode checks on subagents now see the parent session's verified human messages","tier":"notice","area":"Auto Mode","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.299\/e\/subagenthandoff-classifier-now-gets-inherited-verified-huma","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.299","markdown":"### Auto mode checks on subagents now see the parent session's verified human messages\n\nWhen auto mode checks a subagent or a handoff, it is now given the parent session's verified human turns instead of the raw parent messages\n\n**Unclear.** How this changes what auto mode allows or blocks for subagents is not stated.\n\n**What**\n\nAuto mode lets Claude Code decide on its own whether an action may go ahead, using a separate checking model called a classifier. A subagent is a helper Claude starts to work on part of a task, and a handoff passes work from one part of a session to another.\n\nWhen the classifier checks a subagent step or a handoff, it is now given the turns from the parent session that are confirmed to come from a human, along with the earlier part of that conversation. Before, it was handed the parent's messages directly.\n\n**Why**\n\nThe classifier now judges subagent actions with the person's own requests in view, which may change which subagent actions auto mode allows or blocks.\n\n- Area: Auto Mode\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 3\/5\n- Scope: individual\n- Heads-up: no"}