{"version":"2.1.280","anchor":"auto-mode-tool-classification-requires-an-explicit-classifi","canonical_anchor":"auto-mode-tool-classification-requires-an-explicit-classifi","heading":"Auto-mode tool classification tightened with a new 'classifierAllowed' check","tier":"internal","area":"Auto Mode","url":"https:\/\/changelogs.core-directive.com\/v\/2.1.280\/e\/auto-mode-tool-classification-requires-an-explicit-classifi","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.280","markdown":"### Auto-mode tool classification tightened with a new 'classifierAllowed' check\n\nAuto-mode tool classification now tracks an explicit classifierAllowed flag on tool-use records before treating a decision as classifier-driven\n\n**Unclear.** The finding does not say what changes for the user when a call fails this stricter check.\n\n**What**\n\n- Tool-use records now carry a new `autoModeClassified` field (alongside `toolUseId`).\n\n- The `autoModeClassified` marker passed into a tool call now also requires `decisionReason.classifierAllowed === true`, in addition to the existing checks that `decisionReason.type === \"classifier\"` and `classifier === \"auto-mode\"`.\n\n**Why** This adds a stricter, explicit check before a tool call is treated as auto-mode classified, reducing the chance of misclassifying decisions.\n\n- Area: Auto Mode\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5"}