fix: group_gemm format "TN" + Layer 3 ops_api dispatch from xllm upstream

AST chain alignment with upstream_ref/xllm/xllm/core/kernels/ilu/:

Layer 5: ixformer::infer (binary .so on device)
Layer 4: xllm_kernels/ilu/*.cpp -> calls ixformer::infer (0-diff with upstream)
Layer 3: xllm_kernels/ops_api.h+cpp + param.h (NEW from upstream 2719 lines)
         kernels/kernels.h aggregation header (NEW)
Layer 2: xllm_layers/ilu/*.cpp (0-diff with upstream)
Layer 1: ix_full_bridge_v2.cpp pybind11 bridge (FIXED)

Critical fixes in ix_full_bridge_v2.cpp:
- group_gemm format "default" -> "TN" (match upstream ilu/group_gemm.cpp)
- fused_moe_forward: pass 3D weights directly instead of .view({-1,...})
- group_gemm output_n: use tokens_per_experts.sum() per upstream convention

ex_engine/xllm_kernels/kernels.h

@@ -0,0 +1,11 @@
+/* Auto-generated aggregation header for xllm::kernel namespace.
+ * Equivalent to CMake cc_library(NAME kernels HDRS param.h ops_api.h).
+ * 
+ * AST Layer 3: kernel dispatch interface
+ * Called by: xllm_layers/ (Layer 2)
+ * Calls:     xllm_kernels/ilu/ (Layer 4)
+ */
+#pragma once
+
+#include "param.h"
+#include "ops_api.h"