Centralize state-aware telemetry version handling in native execution boundary

Relevant area(s)

Windows, Linux, macOS

Brief description of your issue

State-aware telemetry schema/version selection and rejection are duplicated in the Node and .NET SDKs instead of being owned by the common executor/engine boundary. The duplicated logic can drift across SDKs and makes protocol policy a client responsibility.

Steps to reproduce

1. Inspect sdk/node/src/state-aware-helper.ts telemetry version selection and validation.
2. Inspect sdk/dotnet/Microsoft.Mxc.Sdk/MxcLifecycle.cs ApplyTelemetry and SupportsStableTelemetry.
3. Compare both implementations with native state-aware request parsing and dispatch.

Expected behavior

SDKs serialize caller intent without independently implementing telemetry protocol-version policy. The common native executor/engine boundary selects or validates the applicable contract consistently for every SDK.

Actual behavior

Node and .NET each choose schema versions and reject telemetry on older versions independently, duplicating policy and creating cross-SDK drift risk.

Start by reading sdk/node/src/state-aware-helper.ts and sdk/dotnet/Microsoft.Mxc.Sdk/MxcLifecycle.cs, then compare their telemetry handling with native state-aware request parsing and dispatch. Trace where the common executor or engine boundary can own version selection and rejection. Done means SDKs serialize caller intent without independently enforcing telemetry protocol policy.