fix: accept max and ultra reasoning efforts for task --effort#454

Open

Caleb0796 wants to merge 7 commits into

openai:mainfrom

Caleb0796:fix/accept-max-ultra-reasoning-efforts

Commits on Jul 8, 2026

fix: accept max and ultra reasoning efforts for task --effort

authored and
caleb
committed
fix: validate explicit reasoning efforts against the resolved model's catalog

authored and
caleb wei
committed
fix: validate before thread creation and document per-model tier variance

authored and
caleb wei
committed
fix: validate only explicit model+effort pairs; keep broker stream ownership on request failures

authored and
caleb wei
committed
fix: validate the requested resume model before thread/resume

authored and
caleb wei
committed
fix: refcount broker request ownership; document legacy none/minimal efforts

authored and
caleb wei
committed
fix: fetch the full model catalog (includeHidden + pagination) for effort validation

authored and
caleb wei
committed