Follow Discord
Sweep 29 Sep 2026 · 18:10Z Build v2.1.285 506 read Stable v2.1.280 Latest v2.1.285 Next v2.1.285 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.284 ·

/loop has most models schedule the next run before confirming

When /loop paces itself, most models are now told to decide and schedule the next run first and write their confirmation afterwards

You'll notice Improvements
JSON All of v2.1.284
You'll noticeTier: how much it should matter to you
1Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
Slash CommandsArea: what it touches
ImprovementsKind: in v2.1.284,
ImprovementsSection of the release
What

When /loop runs without a fixed interval, Claude chooses when to run next by setting a wake-up. The order of steps in its instructions now depends on the model:

  • Models with a particular per-model setting turned on, which is the default for claude-mythos-5, keep the old order: confirm briefly, then set the wake-up.
  • Other models are told to decide whether the loop continues, set the wake-up, and only then confirm briefly.

The wording for setting the wake-up again after a task notification is reordered in the same way. Before, every model was told to write its confirmation before setting the wake-up.

Why

Claude's turn ends as soon as the wake-up is set. Setting it before writing the status text probably makes it less likely that a loop stops because the turn ended before the next run was scheduled.

Read from
Names in the bundle/loop
How sure we are
Something disagreesSomething we can check disagrees with this entry, or the writer said they could not settle it.
The writer flagged doubtIt is not clear which models besides claude-mythos-5 keep the old order.
Anthropic's release notes agreeFixed /loop status updates in self-paced mode often not being shown because Claude wrote them only in its reasoning; Claude now writes each…

See this entry in the whole of v2.1.284 →

Feedback