Currently timeouts aren't handled consistently in the client: 1. Streaming timeout is reset to `None` for all methods except for send streaming message for JSON-RPC. 2. gRPC does not update timeout. References: #684, #475. It should be possible to provide timeout per-request at the client level so that each transport can handle it in implementation specific way.
Activity
refactor(client): map error responses to domain errors (#761)
refactor(client): allow transport agnostic per invocation timeouts (#776
ishymko commentedon Mar 6, 2026
Fixed via #776.