{"version":"2.1.323","anchor":"agent-namer-adds-extrabetas-and-prompt-lead","canonical_anchor":"agent-namer-adds-extrabetas-and-prompt-lead","heading":"The request that names agents now sends extra options and an opening line","tier":"internal","area":"Subagents","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.323\/e\/agent-namer-adds-extrabetas-and-prompt-lead","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.323","markdown":"### The request that names agents now sends extra options and an opening line\n\nWhen Claude Code asks the model for a short name for an agent's job, the request now adds extra API options and an opening line\n\n**What**\n\nClaude Code asks the model for a 2 to 4 word lowercase label that names an agent's job. An agent here is a separate helper that Claude Code runs to carry out part of a task. That request has changed in two ways:\n\n- It now turns on extra optional API features, called betas. Before, it used none.\n\n- The prompt now opens with an extra lead-in. Before, it began directly with the instruction to write the label.\n\n**Why**\n\nThis changes the request Claude Code sends when it names an agent, so the names it produces may come out differently.\n\n- Area: Subagents\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}