Source Intelligence

DisclaimerUnofficial, and not affiliated with Anthropic. Nearly all of this is read straight out of what ships: npm bundles, captured prompts, published docs. Anthropic's own notes go in verbatim, marked as theirs. The rest is my reading, and every entry carries the strings behind it. If one looks wrong, vote it down and say why.

All of v2.1.124 Home All releases olderv2.1.123 v2.1.126newer
Claude Code v2.1.124

doneMeansMerged mode (@internal, gradual rollout)

What

A new internal setting that changes Claude's stop condition: instead of stopping when its response is "done", Claude keeps working until the PR is ready to merge, a cron/Monitor is armed to resume later, or it hands you a concrete next step.

Usage (in settings.json, internal builds only):

{
  "doneMeansMerged": true
}
Details
  • Schema description: @internal When true, Claude keeps working until the PR is ready for you to merge, a cron/Monitor is armed to resume later, or it hands you a self-contained next step.
  • A companion meta-system message is injected when the setting is on: The user has asked you to work without stopping for clarifying questions. When you'd normally pause to check, make the reasonable call and continue; they'll redirect if needed. Internal text describing how the harness signals itself: You ended the turn without calling SendUserMessage.
  • A new "session PR resolved" tracker (sessionPrResolved) and the kairos brief stop-hook text are wired into this mode; a flag-controlled hint (tengu_kairos_brief_stop_hook_text) lets the server adjust the in-prompt language without a release.
Evidence

Schema setting and behavioral hint (search for "doneMeansMerged" and "You ended the turn without calling SendUserMessage.").

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

See this entry in the whole of v2.1.124 →