fix(core): Graceful handling of NFE loops and incremental SARIF flushing

strix/core/execution.py

@@ -284,7 +284,7 @@ async def _run_noninteractive_until_lifecycle(
     result: RunResultBase | None = None
     input_data: Any = initial_input
     invalid_final_outputs = 0
-    invalid_final_output_limit = max(1, max_turns)
+    invalid_final_output_limit = min(5, max(1, max_turns))
 
     while True:
         if coordinator.budget_stopped:
@@ -581,6 +581,16 @@ async def _child_loop() -> None:
             )
         except BudgetExceededError:
             logger.info("child %s stopped after reaching the scan budget limit", child_id)
+        except MaxTurnsExceeded as exc:
+            logger.warning("child %s crashed due to MaxTurnsExceeded: %s", child_id, exc)
+            with contextlib.suppress(Exception):
+                await coordinator.set_status(child_id, "crashed")
+                await _notify_parent_on_crash(coordinator, child_id, "crashed")
+        except Exception:
+            logger.exception("child %s crashed with unhandled exception", child_id)
+            with contextlib.suppress(Exception):
+                await coordinator.set_status(child_id, "crashed")
+                await _notify_parent_on_crash(coordinator, child_id, "crashed")
 
     task_handle = asyncio.create_task(_child_loop(), name=f"agent-{name}-{child_id}")
     await coordinator.attach_runtime(child_id, task=task_handle)

strix/core/runner.py

@@ -11,6 +11,7 @@
 
 from agents import RunConfig
 from agents.sandbox import SandboxRunConfig
+from agents.exceptions import MaxTurnsExceeded
 from openai import RateLimitError
 
 from strix.agents.factory import build_strix_agent, make_child_factory
@@ -391,6 +392,17 @@ async def spawn_child_agent(**kwargs: Any) -> dict[str, Any]:
             with contextlib.suppress(Exception):
                 await coordinator.set_status(root_id, "stopped")
         return None
+    except MaxTurnsExceeded as exc:
+        logger.warning(
+            "Scan %s stopped: MaxTurnsExceeded (%s).",
+            scan_id,
+            exc,
+        )
+        if root_id is not None:
+            await coordinator.cancel_descendants(root_id)
+            with contextlib.suppress(Exception):
+                await coordinator.set_status(root_id, "failed")
+        return None
     except BaseException:
         logger.exception("Strix scan %s failed", scan_id)
         if root_id is not None:

strix/tools/agents_graph/tools.py

@@ -13,6 +13,7 @@
 from agents import RunContextWrapper, function_tool
 
 from strix.core.agents import Status, coordinator_from_context
+from strix.report.state import get_global_report_state
 from strix.skills import validate_requested_skills
 
 
@@ -574,6 +575,9 @@ async def agent_finish(
     )
     await coordinator.set_status(me, "completed")
 
+    if report_state := get_global_report_state():
+        report_state.save_run_data()
+
     return json.dumps(
         {
             "success": True,