fix(browser): race condition in session creation can orphan cloud sessions

Two concurrent threads (e.g. parallel subagents) could both pass the
'task_id in _active_sessions' check, both create cloud sessions via
network calls, and then one would overwrite the other — leaking the
first cloud session.

Add double-check after the lock is re-acquired: if another thread
already created a session while we were doing the network call, use
the existing one instead of orphaning it.

tools/browser_tool.py

@@ -555,6 +555,11 @@ def _get_session_info(task_id: Optional[str] = None) -> Dict[str, str]:
             session_info = provider.create_session(task_id)
     
     with _cleanup_lock:
+        # Double-check: another thread may have created a session while we
+        # were doing the network call. Use the existing one to avoid leaking
+        # orphan cloud sessions.
+        if task_id in _active_sessions:
+            return _active_sessions[task_id]
         _active_sessions[task_id] = session_info
     
     return session_info

toolsets.py

@@ -130,12 +130,6 @@ TOOLSETS = {
         "includes": []
     },
     
-    "messaging": {
-        "description": "Cross-platform messaging: send messages to Telegram, Discord, Slack, SMS, etc.",
-        "tools": ["send_message"],
-        "includes": []
-    },
-    
     "rl": {
         "description": "RL training tools for running reinforcement learning on Tinker-Atropos",
         "tools": [