fix: Deadlock in instances_use when port not yet registered

instances_use held _instances_lock while calling register_instance,
which also acquires the same lock — non-reentrant Lock = hang forever.

- Release lock before calling register_instance (avoids blocking
  other threads during the HTTP health check too)
- Upgrade Lock → RLock as safety net for any other reentrant paths

src/ghydramcp/mixins/base.py

@@ -4,7 +4,7 @@ Provides shared state and utilities for all domain mixins.
 """
 
 import time
-from threading import Lock
+from threading import RLock
 from typing import Any, Dict, Optional
 
 from fastmcp import Context
@@ -25,7 +25,7 @@ class GhydraMixinBase(MCPMixin):
 
     # Shared state across all mixins
     _instances: Dict[int, Dict[str, Any]] = {}
-    _instances_lock = Lock()
+    _instances_lock = RLock()
     _current_port: Optional[int] = None
 
     def __init__(self):

src/ghydramcp/mixins/instances.py

@@ -168,8 +168,9 @@ class InstancesMixin(GhydraMixinBase):
             Confirmation message with instance details
         """
         with self._instances_lock:
-            if port not in self._instances:
+            needs_register = port not in self._instances
-                # Try to register it first
+
+        if needs_register:
             result = self.register_instance(port)
             if "Failed" in result or "Error" in result:
                 return result