{"version":"2.1.246","anchor":"multi-agent-reply-pairing-now-rejects-and-counts-suspicious-","canonical_anchor":"multi-agent-reply-pairing-now-rejects-and-counts-suspicious-","heading":"Multi-agent reply pairing now rejects and counts suspicious matches","tier":"internal","area":"Multi-Agent","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.246\/e\/multi-agent-reply-pairing-now-rejects-and-counts-suspicious-","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.246","markdown":"### Multi-agent reply pairing now rejects and counts suspicious matches\n\nReplies from coordinated agents that can't be trusted to match their question are now rejected.\n\n**What**\n\nWhen results come back from coordinated agents, replies that cannot be trusted to match their question are now vetoed and counted by reason: blocked by the delivery gate, duplicate markers, inconsistent copies of the same record, and malformed sources, alongside totals for marked replies and replies that had a coordinator question.\n\n**Evidence**\n\n`veto_inconsistent_copies`\n\n- Area: Multi-Agent\n- Tier: Under the hood\n- Useful: 3\/5\n- Signal: 4\/5"}