feat(agentic-mtp): add cache metrics capture and eval-only support

Add capture_cache_metrics() to snapshot SGLang cache/usage metrics
before and after benchmark runs for debugging. Add eval-only mode
branch to run_eval instead of replay when EVAL_ONLY=true.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: seungrokj

benchmarks/single_node/agentic/dsv4_fp4_mi355x_sglang_mtp.sh

@@ -169,6 +169,16 @@ echo "Starting SGLang server for MI355X..."
 SERVER_PID=$!
 echo "Server PID: $SERVER_PID"
 
+capture_cache_metrics() {
+    {
+        echo "=== SGLang cache metrics snapshot $(date --iso-8601=seconds) ==="
+        curl -fsS "http://localhost:$SGLANG_BACKEND_PORT/metrics" 2>/dev/null \
+            | grep -E '^(sglang:(cache_hit_rate|cached_tokens_total|prompt_tokens_total|hicache_host_used_tokens|hicache_host_total_tokens|token_usage|num_requests_running|num_requests_waiting))' \
+            || true
+        echo "============================================================"
+    } >> "$SERVER_LOG"
+}
+
 wait_for_server_ready --port "$SGLANG_BACKEND_PORT" --server-log "$SERVER_LOG" --server-pid "$SERVER_PID"
 
 if [ "$USE_SGLANG_ROUTER" = "true" ]; then
@@ -190,9 +200,17 @@ if [ "$USE_SGLANG_ROUTER" = "true" ]; then
     echo "Router PID: $ROUTER_PID"
     wait_for_server_ready --port "$PORT" --server-log "$ROUTER_LOG" --server-pid "$ROUTER_PID"
 fi
-
 # ---- Run benchmark ----------------------------------------------------------
-build_replay_cmd "$RESULT_DIR"
-REPLAY_CMD+=" --server-metrics http://localhost:$SGLANG_BACKEND_PORT/metrics"
 
-run_agentic_replay_and_write_outputs "$RESULT_DIR"
+if [ "${#METRICS_ARGS[@]}" -gt 0 ]; then
+    capture_cache_metrics
+    trap capture_cache_metrics EXIT
+fi
+
+if [ "${EVAL_ONLY}" = "true" ]; then
+    run_eval --port "$PORT"
+else
+    build_replay_cmd "$RESULT_DIR"
+    REPLAY_CMD+=" --server-metrics http://localhost:$SGLANG_BACKEND_PORT/metrics"
+    run_agentic_replay_and_write_outputs "$RESULT_DIR"
+fi