{"version":"2.1.71","anchor":"handoff-classifier-sub-agent-safety-review","canonical_anchor":"handoff-classifier-sub-agent-safety-review","heading":"Handoff Classifier (Sub-Agent Safety Review)","tier":null,"area":null,"scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.71\/e\/handoff-classifier-sub-agent-safety-review","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.71","markdown":"### Handoff Classifier (Sub-Agent Safety Review)\n\n**What**\n\nA new safety classifier that reviews sub-agent output when control is handed back to the main agent, checking for potentially dangerous actions performed by sub-agents.\n\n**Details**\n\n- When a sub-agent completes, its actions are reviewed against security rules before the main agent acts on the output\n- If the classifier flags an issue, the main agent sees: \"SECURITY WARNING: This sub-agent performed actions that may violate security policy\"\n- If the classifier is unavailable, a warning note is shown instead of blocking\n- Works in conjunction with the auto mode classifier\n\n**Evidence**\n\nHandoff classifier (search for `\"Handoff classifier flagged sub-agent output\"`)"}