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 v0.2.47 Home All releases olderv0.2.46 v0.2.48newer
Claude Code v0.2.47

Improved Thinking Mode Control

  • New environment variable MAX_THINKING_TOKENS allows precise control over thinking token allocation
  • Set a specific number of thinking tokens instead of relying on keyword detection
  • Example: export MAX_THINKING_TOKENS=8000

See this entry in the whole of v0.2.47 →