fixes and enable tool_runtime tests

.github/workflows/integration-tests.yml

@@ -24,7 +24,7 @@ jobs:
       matrix:
         # Listing tests manually since some of them currently fail
         # TODO: generate matrix list from tests/integration when fixed
-        test-type: [agents, inference, datasets, inspect, scoring, post_training, providers]
+        test-type: [agents, inference, datasets, inspect, scoring, post_training, providers, tool_runtime]
         client-type: [library, http]
       fail-fast: false # we want to run all tests regardless of failure
 

.gitignore

@@ -6,6 +6,7 @@ dev_requirements.txt
 build
 .DS_Store
 llama_stack/configs/*
+.cursor/
 xcuserdata/
 *.hmap
 .DS_Store

llama_stack/providers/utils/tools/mcp.py

@@ -7,7 +7,14 @@
 from contextlib import asynccontextmanager
 from typing import Any
 
-import exceptiongroup
+try:
+    # for python < 3.11
+    import exceptiongroup
+
+    BaseExceptionGroup = exceptiongroup.BaseExceptionGroup
+except ImportError:
+    pass
+
 import httpx
 from mcp import ClientSession
 from mcp import types as mcp_types
@@ -34,7 +41,7 @@ async def sse_client_wrapper(endpoint: str, headers: dict[str, str]):
                 await session.initialize()
                 yield session
     except BaseException as e:
-        if isinstance(e, exceptiongroup.BaseExceptionGroup):
+        if isinstance(e, BaseExceptionGroup):
             for exc in e.exceptions:
                 if isinstance(exc, httpx.HTTPStatusError) and exc.response.status_code == 401:
                     raise AuthenticationRequiredError(exc) from exc

tests/common/mcp.py

@@ -109,8 +109,31 @@ def make_mcp_server(required_auth_token: str | None = None):
             pass
         time.sleep(0.1)
 
-    yield {"server_url": server_url}
+    try:
+        yield {"server_url": server_url}
+    finally:
+        print("Telling SSE server to exit")
+        server_instance.should_exit = True
+        time.sleep(0.5)
 
-    # Tell server to exit
-    server_instance.should_exit = True
-    server_thread.join(timeout=5)
+        # Force shutdown if still running
+        if server_thread.is_alive():
+            try:
+                if hasattr(server_instance, "servers") and server_instance.servers:
+                    for srv in server_instance.servers:
+                        srv.close()
+
+                # Wait for graceful shutdown
+                server_thread.join(timeout=3)
+                if server_thread.is_alive():
+                    print("Warning: Server thread still alive after shutdown attempt")
+            except Exception as e:
+                print(f"Error during server shutdown: {e}")
+
+        # CRITICAL: Reset SSE global state to prevent event loop contamination
+        # Reset the SSE AppStatus singleton that stores anyio.Event objects
+        from sse_starlette.sse import AppStatus
+
+        AppStatus.should_exit = False
+        AppStatus.should_exit_event = None
+        print("SSE server exited")