Skip to content

dagent 0.9.8

Choose a tag to compare

View all tags
@RobotSe7en RobotSe7en released this 04 Sep 08:54
· 27 commits to main since this release
v0.9.8
dba7a96

Highlights

  • Flatten provider reasoning configuration into explicit reasoning_effort and reasoning_capture fields across the SDK, YAML, API, and WebUI.
  • Map reasoning_effort consistently to Chat Completions reasoning_effort and Responses reasoning.effort.
  • Keep response parsing (reasoning_capture) separate from context replay (ContextPolicy.reasoning_replay).

Breaking change

Replace reasoning={"effort": "high", "capture": "field_and_tags"} with reasoning_effort="high" and reasoning_capture="field_and_tags". The old nested object is rejected without a compatibility alias.

Verification

  • Python: 1119 passed, 3 skipped
  • WebUI: 125 passed
  • Production frontend build passed
  • wheel and sdist metadata checks passed

Known limitation

Reasoning effort support remains dependent on the model served by the private vLLM endpoint.

Assets 2
Loading