{"version":"2.1.299","anchor":"reply-continuation-mismatch-reason-in-a-classifier","canonical_anchor":"reply-continuation-mismatch-reason-in-a-classifier","heading":"Tool call comparison can now flag a reply continuation mismatch","tier":"internal","area":"Elsewhere","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.299\/e\/reply-continuation-mismatch-reason-in-a-classifier","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.299","markdown":"### Tool call comparison can now flag a reply continuation mismatch\n\nA step that compares tool calls can now report a reply continuation mismatch, not only a name or input mismatch\n\n**Unclear.** What a `continues_reply` result changes for a user is not known.\n\n**What**\n\nClaude Code has a step that compares tool calls (requests from Claude to use a tool) and reports when the tool name or input differs. It can now also report `continues_reply`. That happens when the call is marked as continuing a reply and the outcome is anything other than flagged.\n\n**Why**\n\nThis looks like part of how Claude Code judges tool calls when deciding on permissions.\n\n- Area: Elsewhere\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}