diff --git a/inferencex-e2e/docs/agentx-standalone.md b/inferencex-e2e/docs/agentx-standalone.md
index 26284994c2..4f2c7a0873 100644
--- a/inferencex-e2e/docs/agentx-standalone.md
+++ b/inferencex-e2e/docs/agentx-standalone.md
@@ -43,6 +43,14 @@ environment variables override preset defaults. Model, tokenizer, dataset,
 concurrency, and output remain caller inputs; `--tokenizer-trust-remote-code` is
 needed only for tokenizers with custom code.
 
+InferenceX's `python3 -m infx.bench agentic` wrapper defaults to the full,
+uncapped corpus (`semianalysis_cc_traces_weka_062126`). Model prefixes beginning
+with `dsr1`, `qwen3.5`, or `qwen3.8next` explicitly select the
+256k-capped corpus (`semianalysis_cc_traces_weka_062126_256k`). New model families
+use the full corpus unless added to `CAPPED_FAMILIES` in
+`infx/bench/agentic/traces.py`. `WEKA_LOADER_OVERRIDE` can pin either loader
+regardless of model family; it does not configure the standalone `agentx` CLI.
+
 The default random seed is **42**. Keep this seed when reproducing AgentX results,
 or pass `--random-seed 123` to use a different seed.
 
diff --git a/inferencex-e2e/infx/bench/agentic/traces.py b/inferencex-e2e/infx/bench/agentic/traces.py
index 021991e4e3..45a2a885ff 100644
--- a/inferencex-e2e/infx/bench/agentic/traces.py
+++ b/inferencex-e2e/infx/bench/agentic/traces.py
@@ -10,12 +10,12 @@
     UNCAPPED: "semianalysisai/cc-traces-weka-062126",
     CAPPED: "semianalysisai/cc-traces-weka-062126-256k",
 }
-UNCAPPED_FAMILIES = ("dsv4", "glm5.2", "glm5.3", "minimaxm3", "kimik3")
+CAPPED_FAMILIES = ("dsr1", "qwen3.5", "qwen3.8next")
 
 
 def resolve(model_prefix: str, override: str | None) -> tuple[str, str]:
     """``(loader, hf_dataset)``; ``override`` (``WEKA_LOADER_OVERRIDE``) pins either corpus."""
-    loader = override or (UNCAPPED if model_prefix.startswith(UNCAPPED_FAMILIES) else CAPPED)
+    loader = override or (CAPPED if model_prefix.startswith(CAPPED_FAMILIES) else UNCAPPED)
     if loader not in LOADERS:
         raise InputError(f"unknown WEKA_LOADER_OVERRIDE={loader!r}; allowed: {', '.join(LOADERS)}")
     return loader, LOADERS[loader]
diff --git a/inferencex-e2e/infx/tests/bench/test_agentic_command.py b/inferencex-e2e/infx/tests/bench/test_agentic_command.py
index 260809124b..a60d67d2fe 100644
--- a/inferencex-e2e/infx/tests/bench/test_agentic_command.py
+++ b/inferencex-e2e/infx/tests/bench/test_agentic_command.py
@@ -15,7 +15,6 @@
 
 import pytest
 
-from infx.bench.agentic import traces
 from infx.bench.agentic.run import Plan, execute
 from infx.bench.agentic.venv import Runtime
 from infx.bench.env import BenchError, InputError
@@ -426,7 +425,7 @@ def frontend(method: str, path: str) -> tuple[int, object]:
 
     assert result.returncode == 0, result.stderr
     assert _events(tmp_path) == [
-        f"hf download --repo-type dataset {traces.LOADERS[traces.CAPPED]}",
+        "hf download --repo-type dataset semianalysisai/cc-traces-weka-062126",
         "GET /v1/models",
         "GET /v1/chat/completions",
         "replay",
diff --git a/inferencex-e2e/infx/tests/bench/test_agentic_replay.py b/inferencex-e2e/infx/tests/bench/test_agentic_replay.py
index e3f7a1c847..46eb7d2163 100644
--- a/inferencex-e2e/infx/tests/bench/test_agentic_replay.py
+++ b/inferencex-e2e/infx/tests/bench/test_agentic_replay.py
@@ -40,7 +40,7 @@ def test_default_point_leaves_the_replay_policy_to_the_scenario():
         "--model", "deepseek-ai/DeepSeek-V4-Pro", "--tokenizer", "deepseek-ai/DeepSeek-V4-Pro",
         "--concurrency", "8", "--benchmark-duration", "3600", "--tokenizer-trust-remote-code",
         "--output-artifact-dir", "/results/aiperf_artifacts",
-        "--public-dataset", traces.CAPPED,
+        "--public-dataset", "semianalysis_cc_traces_weka_062126",
     ]  # fmt: skip
 
 
@@ -168,16 +168,37 @@ def test_malformed_replay_inputs_are_rejected(overrides, message):
 
 
 @pytest.mark.parametrize(
-    ("prefix", "override", "loader"),
+    ("prefix", "override", "capped"),
     [
-        ("big41flash", None, traces.UNCAPPED),
-        ("other", None, traces.CAPPED),
-        ("other", traces.UNCAPPED, traces.UNCAPPED),
+        ("dsr1", None, True),
+        ("qwen3.5", None, True),
+        ("qwen3.8next", None, True),
+        ("musespark", None, False),
+        ("qwen3.5-variant", None, True),
+        ("dsv4", None, False),
+        ("dsv41flash", None, False),
+        ("glm5.2", None, False),
+        ("glm5.3", None, False),
+        ("minimaxm3", None, False),
+        ("kimik3", None, False),
+        ("new-model", None, False),
+        ("qwen3.9", None, False),
+        ("new-model", "", False),
+        ("qwen3.5", traces.UNCAPPED, False),
+        ("new-model", traces.CAPPED, True),
     ],
 )
-def test_corpus_follows_the_model_family_context_unless_pinned(
-    monkeypatch, prefix, override, loader
-):
-    monkeypatch.setattr(traces, "UNCAPPED_FAMILIES", ("big4",))
+def test_corpus_is_uncapped_unless_explicitly_capped_or_pinned(prefix, override, capped):
+    loader, dataset = (
+        ("semianalysis_cc_traces_weka_062126_256k", "semianalysisai/cc-traces-weka-062126-256k")
+        if capped
+        else ("semianalysis_cc_traces_weka_062126", "semianalysisai/cc-traces-weka-062126")
+    )
+    env = {**BASE_ENV, "MODEL_PREFIX": prefix}
+    if override is not None:
+        env["WEKA_LOADER_OVERRIDE"] = override
+
+    config = ReplayConfig.from_env(env, Path("/results"))
 
-    assert traces.resolve(prefix, override) == (loader, traces.LOADERS[loader])
+    assert (config.loader, config.dataset) == (loader, dataset)
+    assert _value(replay_argv(config, "/venv/bin/agentx"), "--public-dataset") == loader
diff --git a/inferencex-e2e/perf-changelog.yaml b/inferencex-e2e/perf-changelog.yaml
index 5737704c28..c16125ad20 100644
--- a/inferencex-e2e/perf-changelog.yaml
+++ b/inferencex-e2e/perf-changelog.yaml
@@ -10097,3 +10097,18 @@
     - "Move the complete GB300 Kimi-K3 AgentX curve to vLLM nightly-49cb2bbf, use probabilistic DSpark sampling throughout (golden AL 3.36/3.62/3.84 for 4/5/7 drafts), and add low-latency megakernel points."
     - "Pin Dynamo ffbbaaa6 (dynamo-parsers 8.1.0), use Kimi-K3 parsers only in evaluation recipes, and set DYN_TCP_REQUEST_TIMEOUT=120 for the frontend and workers."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/4208
+
+- config-keys:
+    - musespark-fp4-mi355x-sglang-agentic-dflash
+    - musespark-fp4-gb200-sglang-agentic-dflash
+    - musespark-fp4-b200-sglang-agentic-dflash
+    - musespark-fp4-h200-sglang-agentic-dflash
+    - musespark-fp4-gb300-sglang-agentic-dflash
+    - musespark-fp4-b300-sglang-agentic-dflash
+    - musespark-fp4-h100-sglang-agentic-dflash
+  scenario-type:
+    - agentic-coding
+  description:
+    - "Default AgentX trace selection to the full corpus, retaining explicit 256k exceptions for dsr1, qwen3.5, and qwen3.8next."
+    - "Move Muse Spark Open AgentX points from the 256k-capped corpus to semianalysis_cc_traces_weka_062126; preserve explicit loader overrides."
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/4287