rob thijssen
869033d08e
Some checks failed
CI / Format (push) Successful in 13s
CI / CUDA type-check (pull_request) Successful in 1m26s
CI / Format (pull_request) Successful in 1m35s
CI / Clippy (pull_request) Successful in 5m19s
CI / Test (push) Successful in 10m37s
CI / Test (pull_request) Failing after 10m25s
CI / Clippy (push) Failing after 14m38s
CI / CUDA type-check (push) Failing after 16m6s
CI / Build cortex SRPM (push) Has been cancelled
CI / Build neuron SRPM (push) Has been cancelled
CI / Publish cortex to COPR (push) Has been cancelled
CI / Publish neuron to COPR (push) Has been cancelled
CI / Bump version in source (push) Has been cancelled
CI / Build cortex SRPM (pull_request) Has been cancelled
CI / Build neuron SRPM (pull_request) Has been cancelled
CI / Publish cortex to COPR (pull_request) Has been cancelled
CI / Publish neuron to COPR (pull_request) Has been cancelled
CI / Bump version in source (pull_request) Has been cancelled
Hermes Agent (and the wider vLLM-convention client ecosystem) probes /v1/models for flat context-window keys and cannot see helexa's limit.context — it fell back to a hardcoded catalogue guess that only matched by luck. Re-add the flat fields additively, derived from the settled `limit` at serialization time (cortex list_models and the router's federation aggregate), omitted when the window is genuinely unknown. `limit` stays the opencode-oriented source of truth. Flips the old regression guard that asserted max_model_len must not appear — the removal it guarded was based on the wrong assumption that the field had no consumer. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01TczcGF7JSjJs8r15RSSGpx