Follow Discord
Sweep 22 Oct 2026 · 04:32Z Build v2.1.308 527 read Stable v2.1.308 Latest v2.1.308 Next v2.1.308 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.307 ·

Agent tool guidance rewritten in the repeated-file-read reminder prompt

Claude's helper-agent instructions were rewritten in the repeated-read reminder prompt, telling it to do small jobs itself and not spawn when in doubt

You'll notice Improvements
JSON All of v2.1.307
You'll noticeTier: how much it should matter to you
1Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
SubagentsArea: what it touches
ImprovementsKind: in v2.1.307,
ImprovementsSection of the release
What

The description of the Agent tool was rewritten in the prompt Claude Code sends with its reminder about reading a file that was already read. The Agent tool is how Claude passes work to a subagent, a separate helper copy of Claude.

The new text sets out when not to delegate:

  • Do the work directly when it is a handful of tool calls (individual actions such as reading a file) or a lookup whose target is already known.
  • Delegate a review only to get an opinion not anchored on Claude's own, and then hand over the code rather than Claude's conclusion.
  • Once something is delegated, wait for the result instead of also doing it.
  • When in doubt, don't spawn a helper.
Why

This text is part of Claude's own instructions, so the rewrite changes how readily Claude starts helpers. The new wording leans toward Claude doing small tasks itself and avoiding duplicated work.

See this entry in the whole of v2.1.307 →

Feedback