diff --git a/README.md b/README.md
index 4077be1..9020f61 100644
--- a/README.md
+++ b/README.md
@@ -33,7 +33,7 @@ Configure the application by setting the following environment variables or upda
 Start the OpenAI-compatible API server:
 
 ```bash
-python serve_mcphost_openai_compatible.py
+python serve_mcphost.py
 ```
 
 Or using uvicorn directly:
diff --git a/serve_mcphost_openai_compatible.py b/serve_mcphost.py
similarity index 69%
rename from serve_mcphost_openai_compatible.py
rename to serve_mcphost.py
index 601f77e..bd5e7f5 100644
--- a/serve_mcphost_openai_compatible.py
+++ b/serve_mcphost.py
@@ -1,3 +1,6 @@
+import asyncio
+import socket
+import subprocess
 import time
 from contextlib import asynccontextmanager
 from typing import List
@@ -109,18 +112,60 @@ async def chat_completions(request: ChatCompletionRequest):
         )
 
 
-# Optional: Add a root endpoint that redirects to documentation
 @app.get("/")
 async def root():
     return {"message": "MCPHost OpenAI-compatible API server. Visit /docs for documentation."}
 
 
-# Optional: Add a health check endpoint
 @app.get("/health")
 async def health_check():
     return {"status": "healthy", "mcphost_alive": mcp_manager._is_alive()}
 
 
+@app.post("/restart-service")
+async def restart_service():
+    """Restart the llm-api-mcphost service using supervisorctl"""
+    hostname = socket.gethostname()
+
+    if hostname == "afsar":
+        async def delayed_restart():
+            await asyncio.sleep(1)  # Wait 1 second
+            try:
+                result = await asyncio.create_subprocess_exec(
+                    "sudo", "supervisorctl", "restart", "llm-api-mcphost",
+                    stdout=asyncio.subprocess.PIPE,
+                    stderr=asyncio.subprocess.PIPE
+                )
+                stdout, stderr = await result.communicate()
+
+                if result.returncode != 0:
+                    logger.error(f"Failed to restart service: {stderr.decode()}")
+                else:
+                    logger.info("Successfully restarted llm-api-mcphost service")
+            except Exception as e:
+                logger.error(f"Error restarting service: {str(e)}")
+
+        # Schedule the restart without waiting for it
+        asyncio.create_task(delayed_restart())
+
+        logger.info("Service restart scheduled")
+        return {
+            "status": "scheduled",
+            "message": "Service restart has been scheduled",
+            "hostname": hostname,
+            "note": "Service will restart in 1 second"
+        }
+    else:
+        logger.info(f"Restart request ignored - hostname is '{hostname}', not 'afsar'")
+        return {
+            "status": "ignored",
+            "message": "Service restart not performed - incorrect hostname",
+            "hostname": hostname,
+            "required_hostname": "afsar"
+        }
+
+
+
 if __name__ == "__main__":
     import uvicorn
     uvicorn.run(app, host=settings.host, port=settings.port)
diff --git a/test.sh b/test.sh
index 4bb6b7f..c4a9197 100755
--- a/test.sh
+++ b/test.sh
@@ -16,10 +16,10 @@ curl -X 'POST' \
   "messages": [
     {
       "role": "user",
-      "content": "sum all first 10 digits of pi? /no_think"
+      "content": "fetch hackernews headlines and provide me a summary of the top 10"
     }
   ],
   "temperature": 0.7,
   "stream": false,
-  "max_tokens": 2058
+  "max_tokens": 4096
 }'