cua-agent 0.2.7__py3-none-any.whl → 0.2.10__py3-none-any.whl

agent/core/__init__.py

@@ -2,7 +2,7 @@
 
 from .factory import BaseLoop
 from .messages import (
-    BaseMessageManager,
+    StandardMessageManager,
     ImageRetentionConfig,
 )
 from .callbacks import (
@@ -18,7 +18,7 @@ __all__ = [
     "BaseLoop",
     "CallbackManager",
     "CallbackHandler",
-    "BaseMessageManager",
+    "StandardMessageManager",
     "ImageRetentionConfig",
     "BaseCallbackManager",
     "ContentCallback",

agent/core/base.py

@@ -5,7 +5,6 @@ import asyncio
 from abc import ABC, abstractmethod
 from typing import Any, AsyncGenerator, Dict, List, Optional
 
-from agent.providers.omni.parser import ParseResult
 from computer import Computer
 from .messages import StandardMessageManager, ImageRetentionConfig
 from .types import AgentResponse
@@ -207,7 +206,7 @@ class BaseLoop(ABC):
     # EVENT HOOKS / CALLBACKS
     ###########################################
     
-    async def handle_screenshot(self, screenshot_base64: str, action_type: str = "", parsed_screen: Optional[ParseResult] = None) -> None:
+    async def handle_screenshot(self, screenshot_base64: str, action_type: str = "", parsed_screen: Optional[dict] = None) -> None:
         """Process a screenshot through callback managers
         
         Args:

agent/core/callbacks.py

@@ -6,8 +6,6 @@ from abc import ABC, abstractmethod
 from datetime import datetime
 from typing import Any, Dict, List, Optional, Protocol
 
-from agent.providers.omni.parser import ParseResult
-
 logger = logging.getLogger(__name__)
 
 class ContentCallback(Protocol):
@@ -117,7 +115,7 @@ class CallbackManager:
         for handler in self.handlers:
             await handler.on_error(error, **kwargs)
             
-    async def on_screenshot(self, screenshot_base64: str, action_type: str = "", parsed_screen: Optional[ParseResult] = None) -> None:
+    async def on_screenshot(self, screenshot_base64: str, action_type: str = "", parsed_screen: Optional[dict] = None) -> None:
         """Called when a screenshot is taken.
         
         Args:
@@ -166,7 +164,7 @@ class CallbackHandler(ABC):
         pass
         
     @abstractmethod
-    async def on_screenshot(self, screenshot_base64: str, action_type: str = "", parsed_screen: Optional[ParseResult] = None) -> None:
+    async def on_screenshot(self, screenshot_base64: str, action_type: str = "", parsed_screen: Optional[dict] = None) -> None:
         """Called when a screenshot is taken.
         
         Args:

agent/core/messages.py

@@ -5,7 +5,6 @@ import json
 from typing import Any, Dict, List, Optional, Union, Tuple
 from dataclasses import dataclass
 import re
-from ..providers.omni.parser import ParseResult
 
 logger = logging.getLogger(__name__)
 
@@ -22,106 +21,6 @@ class ImageRetentionConfig:
         """Check if image retention is enabled."""
         return self.num_images_to_keep is not None and self.num_images_to_keep > 0
 
-
-class BaseMessageManager:
-    """Base class for message preparation and management."""
-
-    def __init__(self, image_retention_config: Optional[ImageRetentionConfig] = None):
-        """Initialize the message manager.
-
-        Args:
-            image_retention_config: Configuration for image retention
-        """
-        self.image_retention_config = image_retention_config or ImageRetentionConfig()
-        if self.image_retention_config.min_removal_threshold < 1:
-            raise ValueError("min_removal_threshold must be at least 1")
-
-        # Track provider for message formatting
-        self.provider = "openai"  # Default provider
-
-    def set_provider(self, provider: str) -> None:
-        """Set the current provider to format messages for.
-
-        Args:
-            provider: Provider name (e.g., 'openai', 'anthropic')
-        """
-        self.provider = provider.lower()
-
-    def prepare_messages(self, messages: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
-        """Prepare messages by applying image retention and caching as configured.
-
-        Args:
-            messages: List of messages to prepare
-
-        Returns:
-            Prepared messages
-        """
-        if self.image_retention_config.should_retain_images():
-            self._filter_images(messages)
-        if self.image_retention_config.enable_caching:
-            self._inject_caching(messages)
-        return messages
-
-    def _filter_images(self, messages: List[Dict[str, Any]]) -> None:
-        """Filter messages to retain only the specified number of most recent images.
-
-        Args:
-            messages: Messages to filter
-        """
-        # Find all tool result blocks that contain images
-        tool_results = [
-            item
-            for message in messages
-            for item in (message["content"] if isinstance(message["content"], list) else [])
-            if isinstance(item, dict) and item.get("type") == "tool_result"
-        ]
-
-        # Count total images
-        total_images = sum(
-            1
-            for result in tool_results
-            for content in result.get("content", [])
-            if isinstance(content, dict) and content.get("type") == "image"
-        )
-
-        # Calculate how many images to remove
-        images_to_remove = total_images - (self.image_retention_config.num_images_to_keep or 0)
-        images_to_remove -= images_to_remove % self.image_retention_config.min_removal_threshold
-
-        # Remove oldest images first
-        for result in tool_results:
-            if isinstance(result.get("content"), list):
-                new_content = []
-                for content in result["content"]:
-                    if isinstance(content, dict) and content.get("type") == "image":
-                        if images_to_remove > 0:
-                            images_to_remove -= 1
-                            continue
-                    new_content.append(content)
-                result["content"] = new_content
-
-    def _inject_caching(self, messages: List[Dict[str, Any]]) -> None:
-        """Inject caching control for recent message turns.
-
-        Args:
-            messages: Messages to inject caching into
-        """
-        # Only apply cache_control for Anthropic API, not OpenAI
-        if self.provider != "anthropic":
-            return
-
-        # Default to caching last 3 turns
-        turns_to_cache = 3
-        for message in reversed(messages):
-            if message["role"] == "user" and isinstance(content := message["content"], list):
-                if turns_to_cache:
-                    turns_to_cache -= 1
-                    content[-1]["cache_control"] = {"type": "ephemeral"}
-                else:
-                    content[-1].pop("cache_control", None)
-                    break
-
-
 class StandardMessageManager:
     """Manages messages in a standardized OpenAI format across different providers."""
 
@@ -160,6 +59,7 @@ class StandardMessageManager:
 
     def get_messages(self) -> List[Dict[str, Any]]:
         """Get all messages in standard format.
+        This method applies image retention policy if configured.
 
         Returns:
             List of messages

agent/providers/anthropic/loop.py

@@ -283,8 +283,12 @@ class AnthropicLoop(BaseLoop):
                 # Create new turn directory for this API call
                 self._create_turn_dir()
 
+
+                # Apply image retention policy
+                self.message_manager.messages = messages.copy()
+                prepared_messages = self.message_manager.get_messages()
                 # Convert standard messages to Anthropic format using utility function
-                anthropic_messages, system_content = to_anthropic_format(messages.copy())
+                anthropic_messages, system_content = to_anthropic_format(prepared_messages)
 
                 # Use API handler to make API call with Anthropic format
                 response = await self.api_handler.make_api_call(

agent/providers/anthropic/utils.py

@@ -4,7 +4,6 @@ import logging
 import re
 from typing import Any, Dict, List, Optional, Tuple, cast
 from anthropic.types.beta import BetaMessage
-from ..omni.parser import ParseResult
 from ...core.types import AgentResponse
 from datetime import datetime
 
@@ -188,7 +187,7 @@ def from_anthropic_format(messages: List[Dict[str, Any]]) -> List[Dict[str, Any]
 async def to_agent_response_format(
     response: BetaMessage,
     messages: List[Dict[str, Any]],
-    parsed_screen: Optional[ParseResult] = None,
+    parsed_screen: Optional[dict] = None,
     parser: Optional[Any] = None,
     model: Optional[str] = None,
 ) -> AgentResponse:

agent/providers/openai/loop.py

@@ -276,7 +276,7 @@ class OpenAILoop(BaseLoop):
                 # Call API
                 screen_size = await self.computer.interface.get_screen_size()
                 response = await self.api_handler.send_initial_request(
-                    messages=messages,
+                    messages=self.message_manager.get_messages(), # Apply image retention policy
                     display_width=str(screen_size["width"]),
                     display_height=str(screen_size["height"]),
                     previous_response_id=self.last_response_id,
@@ -397,7 +397,7 @@ class OpenAILoop(BaseLoop):
                         # The API handler will extract this from the message history
                         if isinstance(self.last_response_id, str):
                             response = await self.api_handler.send_computer_call_request(
-                                messages=self.message_manager.messages,
+                                messages=self.message_manager.get_messages(), # Apply image retention policy
                                 display_width=str(screen_size["width"]),
                                 display_height=str(screen_size["height"]),
                                 previous_response_id=self.last_response_id,  # Use instance variable

agent/providers/uitars/clients/oaicompat.py

@@ -6,6 +6,7 @@ from typing import Dict, List, Optional, Any
 import aiohttp
 import re
 from .base import BaseUITarsClient
+import asyncio
 
 logger = logging.getLogger(__name__)
 
@@ -144,7 +145,7 @@ class OAICompatClient(BaseUITarsClient):
                 else:
                     message = {"role": "user", "content": [{"type": "text", "text": item}]}
                 final_messages.append(message)
-
+                
         payload = {
             "model": self.model, 
             "messages": final_messages, 
@@ -192,7 +193,8 @@ class OAICompatClient(BaseUITarsClient):
                     
                     # if 503, then the endpoint is still warming up
                     if response.status == 503:
-                        logger.error(f"Endpoint is still warming up, please try again later")
+                        logger.error(f"Endpoint is still warming up, trying again in 30 seconds...")
+                        await asyncio.sleep(30)
                         raise Exception(f"Endpoint is still warming up: {response_text}")
                     
                     # Try to parse as JSON if the content type is appropriate

agent/ui/gradio/app.py

@@ -41,7 +41,6 @@ from typing import cast
 # Import from agent package
 from agent.core.types import AgentResponse
 from agent.core.callbacks import DefaultCallbackHandler
-from agent.providers.omni.parser import ParseResult
 from computer import Computer
 
 from agent import ComputerAgent, AgentLoop, LLM, LLMProvider
@@ -103,7 +102,7 @@ class GradioChatScreenshotHandler(DefaultCallbackHandler):
         self,
         screenshot_base64: str,
         action_type: str = "",
-        parsed_screen: Optional[ParseResult] = None,
+        parsed_screen: Optional[dict] = None,
     ) -> None:
         """Add screenshot to chatbot when a screenshot is taken and update the annotated image.
 

{cua_agent-0.2.7.dist-info → cua_agent-0.2.10.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cua-agent
-Version: 0.2.7
+Version: 0.2.10
 Summary: CUA (Computer Use) Agent for AI-driven computer interaction
 Author-Email: TryCua <gh@trycua.com>
 Requires-Python: >=3.11

{cua_agent-0.2.7.dist-info → cua_agent-0.2.10.dist-info}/RECORD

@@ -1,11 +1,11 @@
 agent/__init__.py,sha256=guFGtorDBF6R5hVep0Bvci3_sUJfBlcsq9ss5Kwrej8,1484
-agent/core/__init__.py,sha256=7DhJ_6KKooM6uTmDIlumCnd7OFcU67BYIIR1dpIYUB0,506
+agent/core/__init__.py,sha256=3x4XmLSj40-sjUMOtxOuM82RnOQl0I5AwURk5wW_9GE,514
 agent/core/agent.py,sha256=HUfBe7Uam3TObAmf6KH0GDKuNCNunNmmMcuxS7aZg0Q,8332
-agent/core/base.py,sha256=AiSjnBAcHhZIca4KWBP1vQRE3HyikAPkr4Ij9WDevZQ,8374
-agent/core/callbacks.py,sha256=FKAxyajJ-ZJ5SxNXoupNcrm0GYBgjOjJEsStqst0EAk,6453
+agent/core/base.py,sha256=7hD1rosM-JjyruwSplD4-5YO6BaO1a1bD7bjFYGGUrg,8315
+agent/core/callbacks.py,sha256=uAoJo4rHpVf1d8rzEBFdtSud9jRndPLwDoC4U4uYZlw,6386
 agent/core/experiment.py,sha256=Ywj6q3JZFDKicfPuQsDl0vSN55HS7-Cnk3u3EcUCKe8,8866
 agent/core/factory.py,sha256=zzlCdibctqhf8Uta-SrvE-G7h59wAw-7SGhHiGvS9GY,4608
-agent/core/messages.py,sha256=-OVMDqcxK5MUHPEkHliK29XFJYMRAc1keFvzrUyrOmM,16231
+agent/core/messages.py,sha256=OfwelngzxBxwbbWQWcyZ4ViRdhSQ2YbqqWixeqtbePk,12473
 agent/core/provider_config.py,sha256=jB3fLsEsf806HQZ8jtzfSq4bCYGYONBeuCOoog_Nv_Y,768
 agent/core/telemetry.py,sha256=HElPd32k_w2SJ6t-Cc3j_2-AKdLbFwh2YlM8QViDgRw,4790
 agent/core/tools.py,sha256=53aPme3O8U91n122Smu3TGbyGjQQe2zDimaZgKkFNi0,878
@@ -25,7 +25,7 @@ agent/providers/anthropic/api/logging.py,sha256=vHpwkIyOZdkSTVIH4ycbBPd4a_rzhP7O
 agent/providers/anthropic/api_handler.py,sha256=pWXcqDs0ruviDhRNRrz5Ac9ZH4yDv6ZlwpeG3a42cDg,5206
 agent/providers/anthropic/callbacks/__init__.py,sha256=PciBb6Z6MKSwfXqDjU3pV_0FS4MOn_Np_A7_skD-6dA,104
 agent/providers/anthropic/callbacks/manager.py,sha256=euIah5yiM8nhisN-RWXewo6v0WQr0c-FbMBO04r6dJk,1865
-agent/providers/anthropic/loop.py,sha256=Sepfo0b0oQT98xd3Sv2S7Xc81bfU7L4_Zv3VTiapKkg,21661
+agent/providers/anthropic/loop.py,sha256=T2Ip6Nixsmk67uO-tHIsayrZsMksMp951lnP0QZ6VcM,21847
 agent/providers/anthropic/prompts.py,sha256=EaqyvUb90yybv75VsBYzu4sroga7eMACys0uH9mIVWM,1993
 agent/providers/anthropic/response_handler.py,sha256=ZTprV4NTP9Eb9jQ7QgEKZBX0L6rMj5nqBRiE3Zfws8I,8008
 agent/providers/anthropic/tools/__init__.py,sha256=JyZwuVtPUnZwRSZBSCdQv9yxbLCsygm3l8Ywjjt9qTQ,661
@@ -37,7 +37,7 @@ agent/providers/anthropic/tools/edit.py,sha256=EGRP61MDA4Oue1D7Q-_vLpd6LdGbdBA1Z
 agent/providers/anthropic/tools/manager.py,sha256=yNvgTkfEqnOz5isDF0RxvmBMZB0uh2PipFEH-PUXpoY,2020
 agent/providers/anthropic/tools/run.py,sha256=xhXdnBK1di9muaO44CEirL9hpGy3NmKbjfMpyeVmn8Y,1595
 agent/providers/anthropic/types.py,sha256=SF00kOMC1ui8j9Ah56KaeiR2cL394qCHjFIsBpXxt5w,421
-agent/providers/anthropic/utils.py,sha256=qDp0bFGQhK1dG9U461iaeCiyoVUsksXmD43g9cedRW8,14367
+agent/providers/anthropic/utils.py,sha256=6-lANH2-PjnYcZ_n8uGPbkbk9pqIUad5wh07zzslz3Q,14322
 agent/providers/omni/__init__.py,sha256=5ix67iJdtQNGuGJEjEOF65PwFWO7vdo1QlXD28bRbW4,179
 agent/providers/omni/api_handler.py,sha256=7CpD43lYAqTyNKWfrD8XcM9ekbajqKCTH9p0TWtEQyg,1163
 agent/providers/omni/clients/anthropic.py,sha256=nC_lj3UwrLqx9TIew58yxLqKwrH1_LwJD6EqVSEfp3g,3670
@@ -58,7 +58,7 @@ agent/providers/omni/tools/manager.py,sha256=UhtasaxGcmkxtz-bP1UJ1a4xdYnD3Cv8Pbt
 agent/providers/omni/utils.py,sha256=Ikp6ONL1HO637o3KDtv5yv6q-4uIWAzMSQDvGetWXC8,8724
 agent/providers/openai/__init__.py,sha256=8DS6YNZp42NLCacwXsfRaghyczaOCVovX8TgzXUZf_o,165
 agent/providers/openai/api_handler.py,sha256=L1K56dR1j4JsX1sX4OFYeKoCUMM25Fwj2y9nqv8oOhw,17736
-agent/providers/openai/loop.py,sha256=KWN1I8_t7UGsMiz8Jn9oifPp06aFMFcYaoxBlHb63FA,20622
+agent/providers/openai/loop.py,sha256=_MyjPu4rpHpTxS2nTSRLHrCbSDkZPK5WEG1APKGP-1U,20717
 agent/providers/openai/response_handler.py,sha256=K8v_92uSr9R74Y5INY4naeEZZZm35CLIl4h74MBZhsw,7953
 agent/providers/openai/tools/__init__.py,sha256=-KbHMWcd2OVTk5RYQ3ACBEMygwbH-VW6n_98p0lwM4A,344
 agent/providers/openai/tools/base.py,sha256=Np_BC9Cm6TslK99etE9hVTtsBlcEaGhoNCK3NXdB_Lw,2474
@@ -69,7 +69,7 @@ agent/providers/openai/utils.py,sha256=YeCZWIqOFSeugWoqAS0rhxOKAfL-9uN9nrYSBGBgP
 agent/providers/uitars/__init__.py,sha256=sq5OMVJP9E_sok9tIiKJreGkjmNWXPMObjPTClYv1es,38
 agent/providers/uitars/clients/base.py,sha256=5w8Ajmq1JiPyUQJUAq1lSkfpA8_Ts80NQiDxPMTtQrI,948
 agent/providers/uitars/clients/mlxvlm.py,sha256=lMnN6ecMmWHf_l7khJ2iJHHvT7PE4XagUjrWhB0zEhc,10893
-agent/providers/uitars/clients/oaicompat.py,sha256=uYjwrGCVpFi8wj4kcaJ905ABiY6ksJZXaLlM61B2DUA,8907
+agent/providers/uitars/clients/oaicompat.py,sha256=Aw-HMVqRmgNDw-8UxpXU8td4tvTN4ASqGVaNDGPKhSc,8993
 agent/providers/uitars/loop.py,sha256=m2T7OKHN4HgUO0CLEKF-DD4lCULk_L1MSL4DA8Unt8o,26663
 agent/providers/uitars/prompts.py,sha256=_pQNd438mFpZKZT0aMl6Bd0_GgQxuy9y08kQAMPi9UM,2536
 agent/providers/uitars/tools/__init__.py,sha256=0hc3W6u5TvcXYztYKIyve_C2G3XMfwt_y7grmH0ZHC0,29
@@ -79,8 +79,8 @@ agent/providers/uitars/utils.py,sha256=493STTEEJcVhVbQgR0e8rNTI1DjkxUx8IgIv3wkJ1
 agent/telemetry.py,sha256=pVGxbj0ewnvq4EGj28CydN4a1iOfvZR_XKL3vIOqhOM,390
 agent/ui/__init__.py,sha256=ohhxJLBin6k1hl5sKcmBST8mgh23WXgAXz3pN4f470E,45
 agent/ui/gradio/__init__.py,sha256=ANKZhv1HqsLheWbLVBlyRQ7Q5qGeXuPi5jDs8vu-ZMo,579
-agent/ui/gradio/app.py,sha256=StBehGfPJhE6ywnxU3CHDPkZrOm_2XMT1Npepf89G5c,70675
-cua_agent-0.2.7.dist-info/METADATA,sha256=Jdz7v8P_JvHbN1vEcIyDzf2a51FUJQ5D4WJqL9cLbyA,12688
-cua_agent-0.2.7.dist-info/WHEEL,sha256=tSfRZzRHthuv7vxpI4aehrdN9scLjk-dCJkPLzkHxGg,90
-cua_agent-0.2.7.dist-info/entry_points.txt,sha256=6OYgBcLyFCUgeqLgnvMyOJxPCWzgy7se4rLPKtNonMs,34
-cua_agent-0.2.7.dist-info/RECORD,,
+agent/ui/gradio/app.py,sha256=8NGwdng57sAQ2i2vD2THvpWKdOD-Y7BNYeQGAcj3xtA,70616
+cua_agent-0.2.10.dist-info/METADATA,sha256=UVZuW6ZdH47YKe0Cx5Hd_-RdINufw-HnvsDGw6uXe1A,12689
+cua_agent-0.2.10.dist-info/WHEEL,sha256=tSfRZzRHthuv7vxpI4aehrdN9scLjk-dCJkPLzkHxGg,90
+cua_agent-0.2.10.dist-info/entry_points.txt,sha256=6OYgBcLyFCUgeqLgnvMyOJxPCWzgy7se4rLPKtNonMs,34
+cua_agent-0.2.10.dist-info/RECORD,,