Claude Code 2.1.166 Adds Fallback Models, Glob Deny Rules, and Thinking-Token Controls

Releasebot·high signal

Claude Code 2.1.166 (first seen June 6, 2026) introduces a fallback-models setting that lets you configure up to three models tried in order when the primary is overloaded or unavailable, plus glob pattern support in permission deny rules and explicit thinking-token controls. The release also hardens cross-session messaging and improves JetBrains terminal rendering. This is the latest in a fast June cadence (2.1.160 through 2.1.166 all shipped June 4-6).