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 v2.0.35 Home All releases olderv2.0.34 v2.0.36newer
Claude Code v2.0.35

Fixed Assistant Message Handling

What: Improved reliability of assistant message extraction from streaming responses.

Details:

  • Changed from return on first assistant message to collecting all messages and returning the last one
  • Prevents potential issues where intermediate assistant messages could be missed
  • Adds explicit error if no assistant message found in stream
  • Evidence: ES() at line 468457 (formerly HS())

See this entry in the whole of v2.0.35 →