vLLM issue #6073 request for a dedicated /ready endpoint

Cited by the llm-d readiness-probes guide (ops-llmd-readiness-probes) as the reason vLLM has no model-aware readiness endpoint and that the guide recommends /v1/models instead. Not independently read for this record; recorded as a pointer only, hence draft status and low confidence. Verified 2026-10-10: Read issue '[Feature]: Add readiness endpoint /ready and return /health earlier (vLLM on Kubernetes)' (https://github.com/vllm-project/vllm/issues/6073); confirms feature request by frittentheke for dedicated /ready endpoint and earlier /health return.