perf(agentx): refresh dsv4-fp4-b300-vllm-agentic-mtp harness

benchmarks/single_node/agentic/dsv4_fp4_b300_vllm_mtp.sh

@@ -93,6 +93,13 @@ if [ "$DP_ATTENTION" = "true" ]; then
     agentic_pip_install --quiet "vllm-router==$VLLM_ROUTER_VERSION"
 fi
 
+# AIPerf automatically scrapes the public endpoint's /metrics URL. That is the
+# vLLM engine for pure TP, but the native router for DP-attention. Explicitly
+# add the engine endpoint so every topology captures vLLM metrics; AIPerf
+# deduplicates it against the automatic endpoint in pure-TP runs.
+export AIPERF_SERVER_METRICS_URLS="http://localhost:${VLLM_BACKEND_PORT}/metrics"
+export AIPERF_REQUIRED_SERVER_METRIC_PREFIX="vllm:"
+
 # Match the environment used by v4pro-b300.yaml.
 export VLLM_USE_V2_MODEL_RUNNER=1
 export VLLM_ENGINE_READY_TIMEOUT_S=3600

perf-changelog.yaml

@@ -5526,6 +5526,15 @@
     - "Enable SGLang backend metrics on every aggregate, prefill, and decode engine, and fail before publishing a partial trace artifact if required sglang: metrics are absent."
     - "Record the recipes' active HiCache host-DRAM tier and, for disaggregated points, Dynamo router commit 5a638087 in nvidia-master metadata so generated artifacts no longer report kv_offloading=none, allocated_cpu_dram_gb=0, or a null router."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2477
+
+- config-keys:
+    - dsv4-fp4-b300-vllm-agentic-mtp
+  scenario-type:
+    - agentic-coding
+  description:
+    - "Refresh submission with up to date AgentX harness."
+    - "Capture the vLLM backend Prometheus endpoint for every topology and require vllm: metrics before publishing AgentX results."
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2517
 
 - config-keys:
     - kimik3-fp4-h200-vllm-agentic