{"version":"2.1.301","anchor":"structured-output-prompt-fixed-and-extended","canonical_anchor":"structured-output-prompt-fixed-and-extended","heading":"Structured-output instruction now names the right tool and has extra text","tier":"notice","area":"Elsewhere","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.301\/e\/structured-output-prompt-fixed-and-extended","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.301","markdown":"### Structured-output instruction now names the right tool and has extra text\n\nThe instruction telling Claude to end its turn after a structured-output tool call now names the correct tool and has more text at the end\n\n**What**\n\nClaude Code has an instruction that tells the model to end its turn, with no acknowledgment, after it has called a particular tool to return structured output. That instruction previously filled in the wrong tool name. It now names the correct tool, and a further block of text is added at the end.\n\n**Why**\n\nThis is a correction to wording the model sees. It may make the model more reliable about stopping once its structured answer has been sent.\n\n- Area: Elsewhere\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}