{"version":"2.1.307","anchor":"the-agent-tool-description-in-the-reminder-read-empty-prompt","canonical_anchor":"the-agent-tool-description-in-the-reminder-read-empty-prompt","heading":"Agent tool guidance rewritten in the empty-file reminder prompt","tier":"notice","area":"Subagents","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.307\/e\/the-agent-tool-description-in-the-reminder-read-empty-prompt","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.307","markdown":"### Agent tool guidance rewritten in the empty-file reminder prompt\n\nIn the empty-file reminder prompt, Claude's helper-agent instructions now say to do short tasks and known lookups itself\n\n**What**\n\nThe description of the `Agent` tool was rewritten in the prompt Claude Code sends with its reminder about reading an empty file. The `Agent` tool is how Claude passes work to a subagent, a separate helper copy of Claude.\n\nThe new text tells Claude to do the work itself when it is a handful of tool calls (individual actions such as reading a file) or a lookup whose target it already knows, and not to delegate a check it could run directly.\n\n**Why**\n\nThis text is part of Claude's own instructions, so the rewrite changes when Claude hands work off. Small, simple checks are now meant to stay with Claude rather than go to a helper.\n\n- Area: Subagents\n- Tier: You'll notice\n- Useful: 1\/5\n- Signal: 1\/5\n- Scope: individual\n- Heads-up: no"}