Follow Discord
Sweep 25 Sep 2026 · 19:33Z Build v2.1.283 504 read Stable v2.1.274 Latest v2.1.283 Next v2.1.283 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.280 ·

Thinking mode and display changes are now tracked as their own telemetry reason

Switching thinking mode or display is now tracked end-to-end as a telemetry field and a cache-invalidation/diff reason

Group of 6 Under the hood Internal Changes
JSON All of v2.1.280
Under the hoodTier: how much it should matter to you
1Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
TelemetryArea: what it touches
Internal ChangesKind: in v2.1.280,
Internal ChangesSection of the release

What

  • Telemetry/event schemas gained thinkingMode and thinkingDisplay string fields, each validated against a fixed set of allowed values and defaulting to an empty string.
  • The tracker that compares successive API request configurations now also flags thinkingModeChanged and thinkingDisplayChanged, alongside existing fields like effortChanged and extraBodyChanged.
  • The request-cache invalidation-reason table gained thinking_mode_changed and thinking_display_changed entries, explaining why a cached prompt/request was dropped and re-sent.
  • The turn-configuration-diff logger now reports "thinking mode changed" / "thinking display changed" (with previous/new values) alongside other turn changes like effort or model.
  • The general status-diff helper used to compare snapshots also flags these two changes, but only once a non-empty thinking mode was already set on both sides.

Why

This lets Claude Code correctly explain and log why a request was rebuilt or a cached prompt was invalidated when you change thinking mode or thinking display, instead of leaving those changes unaccounted for in diagnostics and telemetry.

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 doubtThe finding does not say what the allowed values are or which thinking modes/displays they correspond to.

See this entry in the whole of v2.1.280 →

Feedback