diff --git a/src/agentunit/adapters/agentops_adapter.py b/src/agentunit/adapters/agentops_adapter.py
index f02d2b4..768fda6 100644
--- a/src/agentunit/adapters/agentops_adapter.py
+++ b/src/agentunit/adapters/agentops_adapter.py
@@ -67,13 +67,15 @@ def __init__(
         self.project_id = project_id
         self.default_tags = default_tags or []
         self.auto_start_session = auto_start_session
+        self.enable_tracing = kwargs.get("enable_tracing", True)
+        self.client: Any = None
         self.platform = MonitoringPlatform.AGENTOPS
         """
         Initialize LangSmith adapter.
 
         Args:
             api_key: LangSmith API key
-            project_name: Project name for organizing traces
+            project_id: Langsmith project ID
             endpoint: Optional custom LangSmith endpoint
             enable_tracing: Whether to enable automatic tracing
             enable_feedback: Whether to collect feedback data
@@ -98,9 +100,10 @@ def _initialize_agentops(self):
         """Initialize AgentOps client and verify connection."""
         try:
             # Import AgentOps SDK
-            import agentops
+            import agentops  # type: ignore[import-not-found]
 
             self.agentops = agentops
+            self.client = agentops
 
             # Initialize AgentOps
             if self.api_key:
@@ -124,11 +127,6 @@ def _initialize_agentops(self):
             logger.error(f"Failed to connect to AgentOps: {e}")
             raise
 
-    @property
-    def platform(self) -> MonitoringPlatform:
-        """Return the monitoring platform type."""
-        return MonitoringPlatform.AGENTOPS
-
     def create_agent(self, role: AgentRole, agent_id: str | None = None, **kwargs) -> AgentMetadata:
         """
         Create an agent for AgentOps monitoring.
@@ -244,7 +242,7 @@ def send_message(
         interaction = AgentInteraction(
             interaction_id=interaction_id,
             from_agent=from_agent,
-            to_agent=to_agent,
+            to_agent=to_agent or "broadcast",
             content=message,
             timestamp=timestamp,
             metadata=metadata or {},
@@ -339,12 +337,15 @@ def calculate_coordination_metrics(self) -> dict[str, float]:
         for interaction in self.session_interactions:
             unique_agents.add(interaction.from_agent)
             if interaction.to_agent:
-                unique_agents.add(interaction.to_agent)
+                if isinstance(interaction.to_agent, list):
+                    unique_agents.update(interaction.to_agent)
+                else:
+                    unique_agents.add(interaction.to_agent)
 
         agent_participation = len(unique_agents)
 
         # Calculate message distribution
-        agent_counts = {}
+        agent_counts: dict[str, int] = {}
         for interaction in self.session_interactions:
             from_agent = interaction.from_agent
             agent_counts[from_agent] = agent_counts.get(from_agent, 0) + 1
@@ -374,7 +375,7 @@ def calculate_coordination_metrics(self) -> dict[str, float]:
             else 0.0,
         }
 
-    def run_scenario(self, scenario: Scenario) -> ScenarioResult:
+    async def run_scenario(self, scenario: Scenario) -> ScenarioResult:
         """
         Run a scenario with LangSmith integration.
 
@@ -390,14 +391,14 @@ def run_scenario(self, scenario: Scenario) -> ScenarioResult:
         scenario_run_id = None
         if self.enable_tracing:
             try:
-                run = self.client.create_run(
+                run_id = self.agentops.start_trace(
                     name=f"Scenario: {scenario.name}",
                     run_type="chain",
-                    project_name=self.project_name,
-                    inputs={"scenario": scenario.name, "description": scenario.description},
+                    project_name=self.project_id,
+                    inputs={"scenario": scenario.name},
                     tags=["agentunit", "scenario"],
                 )
-                scenario_run_id = str(run.id)
+                scenario_run_id = str(run_id)
             except Exception as e:
                 logger.warning(f"Failed to create scenario run: {e}")
 
@@ -483,15 +484,11 @@ def run_scenario(self, scenario: Scenario) -> ScenarioResult:
             # Update LangSmith run with results
             if scenario_run_id and self.enable_tracing:
                 try:
-                    self.client.update_run(
-                        run_id=scenario_run_id,
-                        outputs={
-                            "result": result.passed,
-                            "execution_time": execution_time,
-                            "details": result.details,
-                        },
-                        end_time=datetime.now(timezone.utc),
+                    self.agentops.update_trace_metadata(
+                        trace_id=scenario_run_id,
+                        metadata={"result": result.success_rate, "details": result.to_dict()},
                     )
+                    self.agentops.end_trace(trace_id=scenario_run_id, status_code="SUCCESS")
                 except Exception as e:
                     logger.warning(f"Failed to update scenario run: {e}")
 
@@ -527,11 +524,11 @@ def run_scenario(self, scenario: Scenario) -> ScenarioResult:
             # Update LangSmith run with error
             if scenario_run_id and self.enable_tracing:
                 try:
-                    self.client.update_run(
-                        run_id=scenario_run_id,
-                        outputs={"error": str(e)},
-                        end_time=datetime.now(timezone.utc),
+                    self.agentops.update_trace_metadata(
+                        trace_id=scenario_run_id,
+                        metadata={"result": result.success_rate, "details": result.to_dict()},
                     )
+                    self.agentops.end_trace(trace_id=scenario_run_id, status_code="SUCCESS")
                 except Exception as e:
                     logger.warning(f"Failed to update failed scenario run: {e}")
 
@@ -551,7 +548,7 @@ def collect_metrics(self, scenario: Any, result: Any, **kwargs) -> ProductionMet
         """
         try:
             # Query recent runs from LangSmith
-            runs = list(self.client.list_runs(project_name=self.project_name, limit=100))
+            runs = list(self.client.list_runs(project_name=self.project_id, limit=100))
 
             if not runs:
                 return ProductionMetrics(
@@ -637,7 +634,7 @@ def establish_baseline(
             # Query historical runs
             runs = list(
                 self.client.list_runs(
-                    project_name=self.project_name, start_time=start_date, end_time=end_date
+                    project_name=self.project_id, start_time=start_date, end_time=end_date
                 )
             )
 
@@ -749,7 +746,7 @@ def create_evaluation_dataset(
             logger.error(f"Failed to create LangSmith dataset: {e}")
             raise
 
-    def run_evaluation(self, dataset_id: str, evaluator_function: Any, **kwargs) -> dict[str, Any]:
+    def run_evaluation(self, dataset_id: str, evaluator_function: Any, **kwargs) -> Any:
         """
         Run evaluation on a LangSmith dataset.
 
@@ -767,7 +764,7 @@ def run_evaluation(self, dataset_id: str, evaluator_function: Any, **kwargs) ->
             results = evaluate(
                 evaluator_function,
                 data=dataset_id,
-                project_name=f"{self.project_name}-evaluation",
+                experiment_prefix=f"{self.project_id}-evaluation",
                 **kwargs,
             )
 
diff --git a/src/agentunit/core/scenario.py b/src/agentunit/core/scenario.py
index 70bba2a..3523707 100644
--- a/src/agentunit/core/scenario.py
+++ b/src/agentunit/core/scenario.py
@@ -7,7 +7,7 @@
 import random
 from dataclasses import dataclass, field
 from pathlib import Path
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Any
 
 from agentunit.datasets.registry import resolve_dataset
 
@@ -68,7 +68,7 @@ def from_openai_agents(
 
         adapter = OpenAIAgentsAdapter.from_flow(flow, **options)
         ds = resolve_dataset(dataset)
-        scenario_name = name or getattr(flow, "__name__", "openai-agents-scenario")
+        scenario_name = name or getattr(flow, "__name__", None) or "openai-agents-scenario"
         return cls(name=scenario_name, adapter=adapter, dataset=ds)
 
     @classmethod
@@ -95,7 +95,7 @@ def from_autogen(
         orchestrator: object,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         """
         Create scenario from AutoGen orchestrator.
@@ -113,7 +113,7 @@ def from_haystack(
         pipeline: object,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.haystack import HaystackAdapter
 
@@ -128,7 +128,7 @@ def from_llama_index(
         engine: object,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.llama_index import LlamaIndexAdapter
 
@@ -143,7 +143,7 @@ def from_semantic_kernel(
         invoker: object,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.semantic_kernel import SemanticKernelAdapter
 
@@ -158,7 +158,7 @@ def from_phidata(
         agent: object,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.phidata import PhidataAdapter
 
@@ -173,7 +173,7 @@ def from_promptflow(
         flow: object,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.promptflow import PromptFlowAdapter
 
@@ -188,7 +188,7 @@ def from_openai_swarm(
         swarm: object,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.openai_swarm import OpenAISwarmAdapter
 
@@ -204,7 +204,7 @@ def from_anthropic_bedrock(
         model_id: str,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.anthropic_bedrock import AnthropicBedrockAdapter
 
@@ -219,7 +219,7 @@ def from_mistral_server(
         base_url: str,
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.mistral_server import MistralServerAdapter
 
@@ -239,7 +239,7 @@ def from_rasa_endpoint(
         target: str | Callable[[dict], object],
         dataset: str | DatasetSource | None = None,
         name: str | None = None,
-        **options: object,
+        **options: Any,
     ) -> Scenario:
         from agentunit.adapters.rasa import RasaAdapter
 
diff --git a/src/agentunit/datasets/builtins.py b/src/agentunit/datasets/builtins.py
index 4b10a0f..4d67445 100644
--- a/src/agentunit/datasets/builtins.py
+++ b/src/agentunit/datasets/builtins.py
@@ -2,7 +2,9 @@
 
 from __future__ import annotations
 
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, TypedDict
+
+from typing_extensions import NotRequired
 
 from .base import DatasetCase, DatasetSource
 
@@ -11,7 +13,18 @@
     from collections.abc import Iterable
 
 
-_GAIA_L1_SHOPPING: list[dict[str, object]] = [
+class DatasetRow(TypedDict):
+    """Provides structure for type hints"""
+
+    id: str
+    query: str
+    expected_output: str | None
+    tools: list[str] | None
+    context: list[str] | None
+    metadata: NotRequired[dict[str, object]]
+
+
+_GAIA_L1_SHOPPING: list[DatasetRow] = [
     {
         "id": "gaia-shopping-001",
         "query": "Find the best price for a pack of AA rechargeable batteries with at least 2500mAh capacity.",
@@ -28,7 +41,7 @@
     },
 ]
 
-_SWE_BENCH_LITE: list[dict[str, object]] = [
+_SWE_BENCH_LITE: list[DatasetRow] = [
     {
         "id": "swe-lite-001",
         "query": "Fix the bug where the API returns HTTP 500 when the username is missing.",
@@ -48,7 +61,7 @@
 ]
 
 
-def _build_loader(rows: list[dict[str, object]]) -> Iterable[DatasetCase]:
+def _build_loader(rows: list[DatasetRow]) -> Iterable[DatasetCase]:
     for row in rows:
         yield DatasetCase(
             id=row["id"],
diff --git a/src/agentunit/metrics/builtin.py b/src/agentunit/metrics/builtin.py
index 3369dc4..638d0c7 100644
--- a/src/agentunit/metrics/builtin.py
+++ b/src/agentunit/metrics/builtin.py
@@ -143,8 +143,9 @@ def evaluate(self, case: DatasetCase, trace: TraceLog, outcome: Any) -> MetricRe
         cost = 0.0
 
         # Check trace metadata
-        if trace.metadata and "cost" in trace.metadata:
-            cost = float(trace.metadata["cost"])
+        metadata = getattr(trace, "metadata", {})
+        if metadata and "cost" in metadata:
+            cost = float(metadata["cost"])
 
         # Check outcome
         elif hasattr(outcome, "cost"):
@@ -168,8 +169,9 @@ def evaluate(self, case: DatasetCase, trace: TraceLog, outcome: Any) -> MetricRe
         total_tokens = 0
 
         # Check trace metadata
-        if trace.metadata and "usage" in trace.metadata:
-            usage = trace.metadata["usage"]
+        metadata = getattr(trace, "metadata", {})
+        if metadata and "usage" in metadata:
+            usage = metadata["usage"]
             prompt_tokens = usage.get("prompt_tokens", 0)
             completion_tokens = usage.get("completion_tokens", 0)
             total_tokens = usage.get("total_tokens", 0)
diff --git a/src/agentunit/production/integrations.py b/src/agentunit/production/integrations.py
index f5ab9a7..a7199a1 100644
--- a/src/agentunit/production/integrations.py
+++ b/src/agentunit/production/integrations.py
@@ -97,7 +97,7 @@ def _calculate_baseline_stats(
         self, historical_data: list[dict[str, Any]], metrics: list[str]
     ) -> dict[str, dict[str, dict[str, float]]]:
         """Calculate baseline statistics from historical data."""
-        baseline_stats = {
+        baseline_stats: dict[str, dict[str, dict[str, float]]] = {
             "performance_baseline": {},
             "quality_baseline": {},
             "reliability_baseline": {},