dev: full OpenAI response schema in the VLM shim (#153) #494

Job	Run time
fmt · clippy (pinned)	34s
wasm32 (declarative converters)	47s
test (stable)	3m 9s
check (cuda, ubuntu-latest)	44s
msrv (1.85 core · 1.88 workspace)	34s
check (coreml, macos-latest)	1m 13s
check (directml, windows-latest)	1m 56s
check (cuda,tensorrt, ubuntu-latest)	42s
release & publish	0s
	9m 39s

Provide feedback