{"version":"2.1.292","anchor":"background-classifier-adds-a-leading-ask-addendum-block","canonical_anchor":"background-classifier-adds-a-leading-ask-addendum-block","heading":"Background agent classifier can add a second block of instructions","tier":"internal","area":"Classifier","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.292\/e\/background-classifier-adds-a-leading-ask-addendum-block","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.292","markdown":"### Background agent classifier can add a second block of instructions\n\nThe classifier that labels background agents' work can now add extra instructions, including a variant for project recaps\n\n**What**\n\nClaude Code uses a classifier, a model call that sorts and labels what a background agent is doing. It used to get a single block of instructions. It can now get a second one, either a \"leading ask\" addendum or a variant for project recaps. The usage data it records notes whether that block was added.\n\n**Why**\n\nThis changes the status cards shown for background agents.\n\n- Area: Classifier\n- Tier: Under the hood\n- Useful: 2\/5\n- Signal: 3\/5\n- Scope: individual\n- Heads-up: no"}