POC: Add context scoped attributes with fallback on baggage(DO NOT MERGE)

instrumentation-genai/opentelemetry-instrumentation-fastmcp/pyproject.toml

@@ -20,7 +20,7 @@ classifiers = [
   "Programming Language :: Python :: 3.13",
 ]
 dependencies = [
-  "opentelemetry-api ~= 1.38.0.dev0",
+  "opentelemetry-api ~= 1.41.0.dev0",
   "opentelemetry-instrumentation ~= 0.59b0.dev0",
   "opentelemetry-semantic-conventions ~= 0.59b0.dev0",
   "splunk-otel-util-genai>=0.1.4",

instrumentation-genai/opentelemetry-instrumentation-fastmcp/src/opentelemetry/instrumentation/fastmcp/instrumentation.py

@@ -95,6 +95,18 @@ def _instrument(self, **kwargs):
         meter_provider = kwargs.get("meter_provider")
         logger_provider = kwargs.get("logger_provider")
 
+        if tracer_provider is None:
+            from opentelemetry import trace
+            tracer_provider = trace.get_tracer_provider()
+
+        if meter_provider is None:
+            from opentelemetry import metrics
+            meter_provider = metrics.get_meter_provider()
+
+        if logger_provider is None:
+            from opentelemetry import _logs
+            logger_provider = _logs.get_logger_provider()
+
         # Get the telemetry handler from util-genai
         self._telemetry_handler = get_telemetry_handler(
             tracer_provider=tracer_provider,

instrumentation-genai/opentelemetry-instrumentation-fastmcp/src/opentelemetry/instrumentation/fastmcp/transport_instrumentor.py

@@ -183,19 +183,23 @@ async def traced_handle_request(wrapped, instance, args, kwargs) -> Any:
                         if tracestate:
                             carrier["tracestate"] = tracestate
 
+                        baggage = getattr(request_meta, "baggage", None)
+                        if baggage:
+                            carrier["baggage"] = baggage
+
                         # Also try model_extra for pydantic v2 extra fields
                         if not carrier and hasattr(request_meta, "model_extra"):
                             extra = request_meta.model_extra
                             if extra:
-                                for key in ["traceparent", "tracestate"]:
+                                for key in ["traceparent", "tracestate", "baggage"]:
                                     if key in extra:
                                         carrier[key] = extra[key]
 
                         if carrier:
                             ctx = propagate.extract(carrier)
                             token = context.attach(ctx)
                             _LOGGER.debug(
-                                f"Attached trace context in _handle_request: "
+                                f"Attached trace context and baggage in _handle_request: "
                                 f"carrier={carrier}"
                             )
 

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/main.py

@@ -586,8 +586,8 @@ def main(manual_instrumentation: bool = False) -> None:
     else:
         print("🔑 Using standard OPENAI_API_KEY authentication")
 
-    if manual_instrumentation:
-        _configure_manual_instrumentation()
+    #if manual_instrumentation:
+    _configure_manual_instrumentation()
 
     session_id = str(uuid4())
     user_request = (

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/multi_agent_travel_planner/requirements.txt

@@ -5,9 +5,11 @@ python-dotenv>=1.0.0
 deepeval
 litellm
 
+opentelemetry-exporter-otlp-proto-grpc>=1.38.0
+
 # Splunk OpenTelemetry GenAI packages
-splunk-otel-util-genai==0.1.9
-splunk-otel-util-genai-evals==0.1.7
-splunk-otel-genai-evals-deepeval==0.1.12
-splunk-otel-genai-emitters-splunk==0.1.6
-splunk-otel-instrumentation-langchain==0.1.7
+#splunk-otel-util-genai==0.1.9
+#splunk-otel-util-genai-evals==0.1.7
+#splunk-otel-genai-evals-deepeval==0.1.12
+#splunk-otel-genai-emitters-splunk==0.1.6
+#splunk-otel-instrumentation-langchain==0.1.7

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/sre_incident_copilot/main.py

@@ -639,8 +639,8 @@ def main():
         sys.exit(1)
 
     # Configure manual instrumentation if requested
-    if args.manual_instrumentation:
-        _configure_manual_instrumentation(config)
+    #if args.manual_instrumentation:
+    _configure_manual_instrumentation(config)
 
     # Set up OpenTelemetry environment
     os.environ.setdefault("OTEL_SERVICE_NAME", config.otel_service_name)

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/sre_incident_copilot/mcp_tools/investigation_agent_mcp.py

@@ -40,72 +40,114 @@
 from agents import investigation_agent  # noqa: E402
 from config import Config  # noqa: E402
 
-mcp = FastMCP("investigation-agent")
-
-
-@mcp.tool()
-async def investigate_incident(
-    service_id: str,
-    investigation_checklist: str,
-    scenario_id: str = None,
-) -> dict:
-    """Investigate an incident by querying metrics, logs, and traces.
-
-    This tool exposes the Investigation Agent as an MCP tool that can be called
-    by other agents or external systems.
-
-    Args:
-        service_id: The service identifier to investigate
-        investigation_checklist: JSON string with investigation steps
-        scenario_id: Optional scenario ID for seeded data
-
-    Returns:
-        Dict containing investigation results with hypotheses and evidence
-    """
-    # Create a minimal state for the agent
-    config = Config.from_env()
-    if scenario_id:
-        config.scenario_id = scenario_id
-
-    state = {
-        "service_id": service_id,
-        "scenario_id": scenario_id,
-        "session_id": f"mcp-{asyncio.get_event_loop().time()}",
-        "triage_result": {
-            "investigation_checklist": json.loads(investigation_checklist)
-            if isinstance(investigation_checklist, str)
-            else investigation_checklist,
-        },
-        "current_agent": "investigation",
-        "hypotheses": [],
-        "confidence_score": 0.0,
-        "eval_metrics": {},
-    }
-
-    try:
-        # Run investigation agent
-        updated_state = investigation_agent(state, config)
-
-        # Extract results
-        investigation_result = updated_state.get("investigation_result", {})
-        hypotheses = updated_state.get("hypotheses", [])
-        confidence_score = updated_state.get("confidence_score", 0.0)
-
-        return {
-            "status": "success",
-            "service_id": service_id,
-            "hypotheses": hypotheses,
-            "investigation_result": investigation_result,
-            "confidence_score": confidence_score,
-            "evidence_count": sum(len(h.get("evidence", [])) for h in hypotheses),
-        }
-    except Exception as e:
-        return {
-            "status": "error",
-            "error": str(e),
+
+def _configure_manual_instrumentation():
+    """Configure manual OpenTelemetry instrumentation."""
+
+    from opentelemetry.sdk.trace import TracerProvider
+    from opentelemetry.sdk.trace.export import BatchSpanProcessor
+    from opentelemetry import _events, _logs, metrics, trace
+    from opentelemetry.exporter.otlp.proto.grpc._log_exporter import OTLPLogExporter
+    from opentelemetry.exporter.otlp.proto.grpc.metric_exporter import (
+        OTLPMetricExporter,
+    )
+    from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import (
+        OTLPSpanExporter,
+    )
+    from opentelemetry.sdk._events import EventLoggerProvider
+    from opentelemetry.sdk._logs import LoggerProvider
+    from opentelemetry.sdk._logs.export import BatchLogRecordProcessor
+    from opentelemetry.sdk.metrics import MeterProvider
+    from opentelemetry.sdk.metrics.export import PeriodicExportingMetricReader
+
+    trace.set_tracer_provider(TracerProvider())
+    trace.get_tracer_provider().add_span_processor(
+        BatchSpanProcessor(OTLPSpanExporter())
+    )
+
+    metric_reader = PeriodicExportingMetricReader(OTLPMetricExporter())
+    metrics.set_meter_provider(MeterProvider(metric_readers=[metric_reader]))
+
+    _logs.set_logger_provider(LoggerProvider())
+    _logs.get_logger_provider().add_log_record_processor(
+        BatchLogRecordProcessor(OTLPLogExporter())
+    )
+    _events.set_event_logger_provider(EventLoggerProvider())
+
+    from opentelemetry.instrumentation.langchain import LangchainInstrumentor
+    instrumentor = LangchainInstrumentor()
+    instrumentor.instrument()
+
+    from opentelemetry.instrumentation.fastmcp import FastMCPInstrumentor
+    instrumentor2 = FastMCPInstrumentor()
+    instrumentor2.instrument()
+
+if __name__ == "__main__":
+    _configure_manual_instrumentation()
+    mcp = FastMCP("investigation-agent")
+
+
+    @mcp.tool()
+    async def investigate_incident(
+            service_id: str,
+            investigation_checklist: str,
+            scenario_id: str = None,
+    ) -> dict:
+        """Investigate an incident by querying metrics, logs, and traces.
+
+        This tool exposes the Investigation Agent as an MCP tool that can be called
+        by other agents or external systems.
+
+        Args:
+            service_id: The service identifier to investigate
+            investigation_checklist: JSON string with investigation steps
+            scenario_id: Optional scenario ID for seeded data
+
+        Returns:
+            Dict containing investigation results with hypotheses and evidence
+        """
+        # Create a minimal state for the agent
+        config = Config.from_env()
+        if scenario_id:
+            config.scenario_id = scenario_id
+
+        state = {
             "service_id": service_id,
+            "scenario_id": scenario_id,
+            "session_id": f"mcp-{asyncio.get_event_loop().time()}",
+            "triage_result": {
+                "investigation_checklist": json.loads(investigation_checklist)
+                if isinstance(investigation_checklist, str)
+                else investigation_checklist,
+            },
+            "current_agent": "investigation",
+            "hypotheses": [],
+            "confidence_score": 0.0,
+            "eval_metrics": {},
         }
 
+        try:
+            # Run investigation agent
+            updated_state = investigation_agent(state, config)
+
+            # Extract results
+            investigation_result = updated_state.get("investigation_result", {})
+            hypotheses = updated_state.get("hypotheses", [])
+            confidence_score = updated_state.get("confidence_score", 0.0)
+
+            return {
+                "status": "success",
+                "service_id": service_id,
+                "hypotheses": hypotheses,
+                "investigation_result": investigation_result,
+                "confidence_score": confidence_score,
+                "evidence_count": sum(len(h.get("evidence", [])) for h in hypotheses),
+            }
+        except Exception as e:
+            return {
+                "status": "error",
+                "error": str(e),
+                "service_id": service_id,
+            }
 
-if __name__ == "__main__":
     mcp.run(transport="stdio", show_banner=False)

instrumentation-genai/opentelemetry-instrumentation-langchain/examples/sre_incident_copilot/tools.py

@@ -366,6 +366,12 @@ def notifier(message: str, channel: str = "incidents") -> str:
     }
     return json.dumps(notification, indent=2)
 
+def _configure_manual_instrumentation():
+    """Configure manual OpenTelemetry instrumentation."""
+
+    from opentelemetry.instrumentation.fastmcp import FastMCPInstrumentor
+    instrumentor2 = FastMCPInstrumentor()
+    instrumentor2.instrument()
 
 @tool
 def investigation_agent_mcp(
@@ -383,7 +389,7 @@ def investigation_agent_mcp(
     Returns:
         JSON string with investigation results
     """
-
+    _configure_manual_instrumentation()
     mcp_script_path = os.path.join(
         os.path.dirname(__file__), "mcp_tools", "investigation_agent_mcp.py"
     )

instrumentation-genai/opentelemetry-instrumentation-langchain/pyproject.toml

@@ -24,7 +24,7 @@ classifiers = [
   "Programming Language :: Python :: 3.13",
 ]
 dependencies = [
-  "opentelemetry-api ~= 1.38.0.dev0",
+  "opentelemetry-api ~= 1.41.0.dev0",
   "opentelemetry-instrumentation ~= 0.59b0.dev0",
   "opentelemetry-semantic-conventions ~= 0.59b0.dev0",
   "splunk-otel-util-genai>=0.1.4",