fix(argo): set FOXHUNT_CUDA_ARCH from detected compute cap

ml-backtesting/build.rs reads FOXHUNT_CUDA_ARCH (defaults sm_86),
ml-alpha/build.rs reads CUDA_COMPUTE_CAP. Template only set the
latter, so lobsim cubins compiled for sm_86 on H100 (sm_90) →
CUDA_ERROR_NO_BINARY_FOR_GPU at runtime.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

infra/k8s/argo/alpha-rl-template.yaml

@@ -158,6 +158,7 @@ spec:
 
             # Auto-detect GPU compute capability for all crates' build.rs
             export CUDA_COMPUTE_CAP=$(nvidia-smi --query-gpu=compute_cap --format=csv,noheader | head -1 | tr -d '.')
+            export FOXHUNT_CUDA_ARCH="sm_${CUDA_COMPUTE_CAP}"
             echo "=== GPU arch: sm_${CUDA_COMPUTE_CAP} ==="
 
             # Clear stale build artifacts when arch changes