Follow Discord
Sweep 01 Oct 2026 · 17:27Z Build v2.1.287 508 read Stable v2.1.285 Latest v2.1.287 Next v2.1.287 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.286 ·

Auto mode's safety check stops testing two wordings

The two-wording experiment in auto mode's safety check has ended, and the classifier now always uses a single wording

Under the hood Internal Changes
JSON All of v2.1.286
Under the hoodTier: how much it should matter to you
1Useful: my rating, 1 to 5
2Signal: worth watching, 1 to 5
Auto ModeArea: what it touches
Internal ChangesKind: in v2.1.286,
Internal ChangesSection of the release
What

In auto mode, a safety check called a classifier decides whether each action may run without asking you. Until now, the classifier could be given one of two wordings, so the two could be compared. That choice has been removed, and the classifier now always uses one wording.

The wording in use is also no longer recorded in the usage data Claude Code reports about auto-mode decisions, or in the classifier's results.

Why

A comparison test between the two wordings is over. Everyone now gets the same classifier instructions.

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 stated which of the two wordings was kept.

See this entry in the whole of v2.1.286 →

Feedback