radio-t / super-bot / 27150093505

75%

push

github

add reasoning_effort option for openai reasoning models

new --reasoning-effort flag (env OPENAI_REASONING_EFFORT), applied to the
request only for reasoning models (o-series, gpt-5) and only when non-empty.
Empty keeps the API default, and non-reasoning models stay untouched since
they reject the parameter.

without it gpt-5-mini runs at the default medium effort, where reasoning
tokens share the max_completion_tokens budget and can starve the visible
reply. low effort gives short chat answers room.

3 of 4 new or added lines in 2 files covered. (75.0%)

2710 of 3608 relevant lines covered (75.11%)

54.22 hits per line