Skip to content

v3.32.0

Choose a tag to compare

@wuweiweiwu wuweiweiwu released this 21 Jul 20:12

Minor Changes

  • 30d85e4: Apply the configured heal failure behavior when failed tests cannot be attempted.

Patch Changes

  • 9d2a7ee: Improve reliability for long-running AI agent conversations.
  • 7ee727a: Make run_step faster by allowing eligible work to finish asynchronously.