cua-agent 0.1.30__tar.gz → 0.1.31__tar.gz

{cua_agent-0.1.30 → cua_agent-0.1.31}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cua-agent
-Version: 0.1.30
+Version: 0.1.31
 Summary: CUA (Computer Use) Agent for AI-driven computer interaction
 Author-Email: TryCua <gh@trycua.com>
 Requires-Python: >=3.10
@@ -101,6 +101,7 @@ pip install "cua-agent[all]"
 # or install specific loop providers
 pip install "cua-agent[openai]" # OpenAI Cua Loop
 pip install "cua-agent[anthropic]" # Anthropic Cua Loop
+pip install "cua-agent[uitars]"    # UI-Tars support
 pip install "cua-agent[omni]" # Cua Loop based on OmniParser (includes Ollama for local models)
 pip install "cua-agent[ui]" # Gradio UI for the agent
 ```
@@ -148,7 +149,13 @@ Refer to these notebooks for step-by-step guides on how to use the Computer-Use
 
 ## Using the Gradio UI
 
-The agent includes a Gradio-based user interface for easy interaction. To use it:
+The agent includes a Gradio-based user interface for easier interaction.
+
+<div align="center">
+    <img src="../../img/agent_gradio_ui.png"/>
+</div>
+
+To use it:
 
 ```bash
 # Install with Gradio support

{cua_agent-0.1.30 → cua_agent-0.1.31}/README.md

@@ -31,6 +31,7 @@ pip install "cua-agent[all]"
 # or install specific loop providers
 pip install "cua-agent[openai]" # OpenAI Cua Loop
 pip install "cua-agent[anthropic]" # Anthropic Cua Loop
+pip install "cua-agent[uitars]"    # UI-Tars support
 pip install "cua-agent[omni]" # Cua Loop based on OmniParser (includes Ollama for local models)
 pip install "cua-agent[ui]" # Gradio UI for the agent
 ```
@@ -78,7 +79,13 @@ Refer to these notebooks for step-by-step guides on how to use the Computer-Use
 
 ## Using the Gradio UI
 
-The agent includes a Gradio-based user interface for easy interaction. To use it:
+The agent includes a Gradio-based user interface for easier interaction.
+
+<div align="center">
+    <img src="../../img/agent_gradio_ui.png"/>
+</div>
+
+To use it:
 
 ```bash
 # Install with Gradio support

{cua_agent-0.1.30 → cua_agent-0.1.31}/agent/providers/omni/clients/oaicompat.py

@@ -93,7 +93,14 @@ class OAICompatClient(BaseOmniClient):
         """
         headers = {"Content-Type": "application/json", "Authorization": f"Bearer {self.api_key}"}
 
-        final_messages = [{"role": "system", "content": system}]
+        final_messages = [
+            {
+                "role": "system", 
+                "content": [
+                    { "type": "text", "text": system }
+                ]
+            }
+        ]
 
         # Process messages
         for item in messages:
@@ -117,7 +124,10 @@ class OAICompatClient(BaseOmniClient):
                     else:
                         message = {
                             "role": item["role"],
-                            "content": [{"type": "text", "text": item["content"]}],
+                            "content": [{
+                                "type": "text", 
+                                "text": item["content"]
+                            }],
                         }
                     final_messages.append(message)
             else:

{cua_agent-0.1.30 → cua_agent-0.1.31}/agent/providers/openai/tools/computer.py

@@ -162,8 +162,8 @@ class ComputerTool(BaseComputerTool, BaseOpenAITool):
                 y = kwargs.get("y")
                 if x is None or y is None:
                     raise ToolError("x and y coordinates are required for scroll action")
-                scroll_x = kwargs.get("scroll_x", 0) // 20
-                scroll_y = kwargs.get("scroll_y", 0) // 20
+                scroll_x = kwargs.get("scroll_x", 0) // 50
+                scroll_y = kwargs.get("scroll_y", 0) // 50
                 return await self.handle_scroll(x, y, scroll_x, scroll_y)
             elif type == "screenshot":
                 return await self.screenshot()
@@ -240,11 +240,7 @@ class ComputerTool(BaseComputerTool, BaseOpenAITool):
 
             if len(mapped_keys) > 1:
                 # For key combinations (like Ctrl+C)
-                for k in mapped_keys:
-                    await self.computer.interface.press_key(k)
-                await asyncio.sleep(0.1)
-                for k in reversed(mapped_keys):
-                    await self.computer.interface.press_key(k)
+                await self.computer.interface.hotkey(*mapped_keys)
             else:
                 # Single key press
                 await self.computer.interface.press_key(mapped_keys[0])

{cua_agent-0.1.30 → cua_agent-0.1.31}/agent/providers/uitars/clients/oaicompat.py

@@ -94,8 +94,15 @@ class OAICompatClient(BaseUITarsClient):
         """
         headers = {"Content-Type": "application/json", "Authorization": f"Bearer {self.api_key}"}
 
-        final_messages = [{"role": "system", "content": system}]
-
+        final_messages = [
+            {
+                "role": "system", 
+                "content": [
+                    { "type": "text", "text": system }
+                ]
+            }
+        ]
+        
         # Process messages
         for item in messages:
             if isinstance(item, dict):
@@ -138,8 +145,13 @@ class OAICompatClient(BaseUITarsClient):
                     message = {"role": "user", "content": [{"type": "text", "text": item}]}
                 final_messages.append(message)
 
-        payload = {"model": self.model, "messages": final_messages, "temperature": self.temperature}
-        payload["max_tokens"] = max_tokens or self.max_tokens
+        payload = {
+            "model": self.model, 
+            "messages": final_messages, 
+            "max_tokens": max_tokens or self.max_tokens,
+            "temperature": self.temperature,
+            "top_p": 0.7,
+        }
 
         try:
             async with aiohttp.ClientSession() as session:

{cua_agent-0.1.30 → cua_agent-0.1.31}/agent/providers/uitars/loop.py

@@ -20,7 +20,7 @@ from computer import Computer
 from .utils import add_box_token, parse_actions, parse_action_parameters
 from .tools.manager import ToolManager
 from .tools.computer import ToolResult
-from .prompts import COMPUTER_USE, SYSTEM_PROMPT
+from .prompts import COMPUTER_USE, SYSTEM_PROMPT, MAC_SPECIFIC_NOTES
 
 from .clients.oaicompat import OAICompatClient
 
@@ -184,7 +184,7 @@ class UITARSLoop(BaseLoop):
         if first_user_idx is not None and instruction:
             # Create the computer use prompt
             user_prompt = COMPUTER_USE.format(
-                instruction=instruction,
+                instruction='\n'.join([instruction, MAC_SPECIFIC_NOTES]),
                 language="English"
             )
             
@@ -232,8 +232,11 @@ class UITARSLoop(BaseLoop):
                     if self.client is None:
                         raise RuntimeError("Failed to initialize client")
 
-                # Convert messages to UI-TARS format
+                # Get messages in standard format from the message manager
+                self.message_manager.messages = messages.copy()
                 prepared_messages = self.message_manager.get_messages()
+                
+                # Convert messages to UI-TARS format
                 uitars_messages = self.to_uitars_format(prepared_messages)
                 
                 # Log request

{cua_agent-0.1.30 → cua_agent-0.1.31}/agent/providers/uitars/prompts.py

@@ -1,5 +1,9 @@
 """Prompts for UI-TARS agent."""
 
+MAC_SPECIFIC_NOTES = """
+(You are operating on macOS, use 'cmd' instead of 'ctrl' for most shortcuts e.g., hotkey(key='cmd c') for copy, hotkey(key='cmd v') for paste, hotkey(key='cmd t') for new tab).)
+"""
+
 SYSTEM_PROMPT = "You are a helpful assistant."
 
 COMPUTER_USE = """You are a GUI agent. You are given a task and your action history, with screenshots. You need to perform the next action to complete the task.
@@ -56,4 +60,4 @@ finished(content='xxx') # Use escape characters \\', \\", and \\n in content par
 
 ## User Instruction
 {instruction}
-""" 
+"""

{cua_agent-0.1.30 → cua_agent-0.1.31}/agent/providers/uitars/tools/computer.py

@@ -173,9 +173,13 @@ class ComputerTool(BaseComputerTool):
             elif action == "hotkey":
                 if "keys" in kwargs:
                     keys = kwargs["keys"]
-                    for key in keys:
-                        await self.computer.interface.press_key(key)
                     
+                    if len(keys) > 1:
+                        await self.computer.interface.hotkey(*keys)
+                    else:
+                        # Single key press
+                        await self.computer.interface.press_key(keys[0])
+                        
                     # Wait for UI to update
                     await asyncio.sleep(0.3)
                     

{cua_agent-0.1.30 → cua_agent-0.1.31}/pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "pdm.backend"
 
 [project]
 name = "cua-agent"
-version = "0.1.30"
+version = "0.1.31"
 description = "CUA (Computer Use) Agent for AI-driven computer interaction"
 readme = "README.md"
 authors = [
@@ -108,7 +108,7 @@ target-version = [
 
 [tool.ruff]
 line-length = 100
-target-version = "0.1.30"
+target-version = "0.1.31"
 select = [
     "E",
     "F",
@@ -122,7 +122,7 @@ docstring-code-format = true
 
 [tool.mypy]
 strict = true
-python_version = "0.1.30"
+python_version = "0.1.31"
 ignore_missing_imports = true
 disallow_untyped_defs = true
 check_untyped_defs = true