{"version":"2.1.302","anchor":"auto-mode-classifier-server-side-input-too-long-handled-sep","canonical_anchor":"auto-mode-classifier-server-side-input-too-long-handled-sep","heading":"Auto mode handles the server rejecting an over-long safety check separately","tier":"notice","area":"Auto Mode","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.302\/e\/auto-mode-classifier-server-side-input-too-long-handled-sep","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.302","markdown":"### Auto mode handles the server rejecting an over-long safety check separately\n\nWhen the auto mode safety check is too long and the server rejects it, Claude Code now treats some cases differently and compacts only for some\n\n**Unclear.** Which of these over-long cases now lead to compaction, and whether this handling is switched on, is not settled.\n\n**What**\n\nIn auto mode, a classifier (a separate model call that judges whether an action is safe) checks actions before Claude Code runs them, reading a copy of the conversation. When that copy is too long and the server reports that the input is too long, Claude Code now handles it as its own case. Actions from MCP tools (tools provided by outside servers) and paths that do not compact the conversation are treated differently, and compaction, which shortens the conversation by summarising it, is requested only for some of these cases.\n\n**Why**\n\nThis changes how auto mode copes when a conversation grows too long for its safety check.\n\n- Area: Auto Mode\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}