replayer: restore optional max_inflight_sessions for backwards compat

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

replayer/replay.py

@@ -59,6 +59,7 @@ class ReplayConfig:
     request_timeout_s: float = 600.0
     request_limit: int | None = None
     model_name: str = "default"
+    max_inflight_sessions: int | None = None  # cap on concurrent sessions; None = unlimited
 
 
 def _build_prompt_token_ids(req: TraceRequest) -> list[int]:
@@ -215,7 +216,11 @@ async def _run_session(
     earliest_ts: float,
     sweep_start: float,
     sink: IncrementalMetricSink,
+    session_sem: asyncio.Semaphore | None = None,
 ) -> list[RequestMetrics]:
+    if session_sem is not None:
+        await session_sem.acquire()
+    try:
         for req in state.turns:
             # Wait until this request's trace timestamp
             target_wall = (req.timestamp_s - earliest_ts)
@@ -230,6 +235,9 @@ async def _run_session(
             )
             state.metrics.append(metric)
             await sink.append(metric)
+    finally:
+        if session_sem is not None:
+            session_sem.release()
 
     return state.metrics