René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Bug: RuntimeError during cleanup on Windows – no running event loop","articleBody":"---\n\n### **Describe the bug**  \nWhen running the `mcp` client using `uv run client.py` on **Windows**, the client connects and runs correctly, but after the response is returned, a `RuntimeError: no running event loop` is thrown during shutdown. The error occurs inside the `terminate_windows_process` function in `win32.py`, which uses `anyio.fail_after()` after the event loop has already closed.\n\n---\n\n### **To Reproduce**\nSteps to reproduce the behavior:\n1. Use the following Python code to create a minimal MCP client that connects to a server and runs a tool-enabled Groq chat query:\n\n```python\n# save this as `client.py`\nimport asyncio\nimport json\nfrom contextlib import AsyncExitStack\nimport os\nfrom typing import Any, Dict, List, Optional\n\nimport nest_asyncio\nfrom dotenv import load_dotenv\nfrom mcp import ClientSession, StdioServerParameters\nfrom mcp.client.stdio import stdio_client\nfrom groq import AsyncGroq\n\nnest_asyncio.apply()\nload_dotenv()\n\nclass MCPGroqClient:\n    def __init__(self, model: str = \"llama-3.3-70b-versatile\"):\n        self.session: Optional[ClientSession] = None\n        self.exit_stack = AsyncExitStack()\n        self.groq_client = AsyncGroq(api_key=os.getenv(\"GROQ_API_KEY\"))\n        self.model = model\n        self.stdio: Optional[Any] = None\n        self.write: Optional[Any] = None\n\n    async def connect_to_server(self, server_script_path: str = \"server.py\"):\n        server_params = StdioServerParameters(command=\"python\", args=[server_script_path])\n        stdio_transport = await self.exit_stack.enter_async_context(stdio_client(server_params))\n        self.stdio, self.write = stdio_transport\n        self.session = await self.exit_stack.enter_async_context(ClientSession(self.stdio, self.write))\n        await self.session.initialize()\n        tools_result = await self.session.list_tools()\n        print(\"\\nConnected to server with tools:\")\n        for tool in tools_result.tools:\n            print(f\"  - {tool.name}: {tool.description}\")\n\n    async def get_mcp_tools(self) -\u003e List[Dict[str, Any]]:\n        tools_result = await self.session.list_tools()\n        return [\n            {\n                \"type\": \"function\",\n                \"function\": {\n                    \"name\": tool.name,\n                    \"description\": tool.description,\n                    \"parameters\": tool.inputSchema,\n                },\n            }\n            for tool in tools_result.tools\n        ]\n\n    async def process_query(self, query: str) -\u003e str:\n        tools = await self.get_mcp_tools()\n        response = await self.groq_client.chat.completions.create(\n            model=self.model,\n            messages=[{\"role\": \"user\", \"content\": query}],\n            tools=tools,\n            tool_choice=\"auto\",\n        )\n        assistant_message = response.choices[0].message\n        messages = [{\"role\": \"user\", \"content\": query}, assistant_message]\n\n        if assistant_message.tool_calls:\n            for tool_call in assistant_message.tool_calls:\n                result = await self.session.call_tool(\n                    tool_call.function.name,\n                    arguments=json.loads(tool_call.function.arguments),\n                )\n                messages.append({\n                    \"role\": \"tool\",\n                    \"tool_call_id\": tool_call.id,\n                    \"content\": result.content[0].text,\n                })\n\n            final_response = await self.groq_client.chat.completions.create(\n                model=self.model,\n                messages=messages,\n                tools=tools,\n                tool_choice=\"none\",\n            )\n            return final_response.choices[0].message.content\n\n        return assistant_message.content\n\n    async def cleanup(self):\n        await self.exit_stack.aclose()\n\nasync def main():\n    client = MCPGroqClient()\n    await client.connect_to_server()\n    query = \"What is our company's vacation policy?\"\n    print(f\"\\nQuery: {query}\")\n    response = await client.process_query(query)\n    print(f\"\\nResponse: {response}\")\n\nif __name__ == \"__main__\":\n    asyncio.run(main())\n```\n\n2. Set your environment variable `GROQ_API_KEY`.\n3. Start the server with a valid tool (like `knowledge_base`).\n4. Run the client using:\n   ```bash\n   uv run client.py\n   ```\n5. Observe the traceback at shutdown related to `RuntimeError: no running event loop`.\n\n---\n\n### **Expected behavior**  \nThe client should shut down gracefully **without throwing exceptions** after processing a response.\n\n---\n\n### **Screenshots / Traceback**\n```bash\nException ignored in: \u003casync_generator object stdio_client at 0x000002025AE0EFC0\u003e\nTraceback (most recent call last):\n  File \"C:\\Program Files\\Python312\\Lib\\asyncio\\tasks.py\", line 314, in __step_run_and_handle_result\n    result = coro.send(None)\n             ^^^^^^^^^^^^^^^\nRuntimeError: async generator ignored GeneratorExit\n\nException ignored in: \u003ccoroutine object terminate_windows_process at 0x000002025AF5F3E0\u003e\nTraceback (most recent call last):\n  File \"...\\Lib\\site-packages\\mcp\\client\\stdio\\win32.py\", line 105, in terminate_windows_process\n    with anyio.fail_after(2.0):\n  File \"C:\\Program Files\\Python312\\Lib\\contextlib.py\", line 158, in __exit__\n    self.gen.throw(value)\n  File \"...\\Lib\\site-packages\\anyio\\_core\\_tasks.py\", line 112, in fail_after\n    with get_async_backend().create_cancel_scope(\n  File \"...\\Lib\\site-packages\\anyio\\_backends\\_asyncio.py\", line 456, in __exit__\n    if current_task() is not self._host_task:\n       ^^^^^^^^^^^^^^\nRuntimeError: no running event loop\n```\n\n---\n\n### **Desktop (please complete the following information):**\n- **OS:** Windows 11  \n- **Python Version:** 3.12  \n- **MCP Version:** [insert version]  \n- **AnyIO Version:** [insert version]  \n- **Terminal:** Windows PowerShell  \n- **Command Used:** `uv run client.py`\n\n---\n\n### **Additional context**  \n- This issue seems to be a **Windows-specific async shutdown bug**.\n- Happens when `anyio.fail_after()` is used after the event loop is already closed.\n- Suggest adding a check before using async context tools during shutdown, such as:\n  ```python\n  if asyncio.get_event_loop().is_running():\n      with anyio.fail_after(2.0):\n          ...\n  ```\n- Or use `try/except RuntimeError` to suppress it gracefully.\n\n---\n","author":{"url":"https://github.com/Dadiya-Harsh","@type":"Person","name":"Dadiya-Harsh"},"datePublished":"2025-04-23T18:26:36.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":0},"url":"https://github.com/575/python-sdk/issues/575"}