cua-agent 0.1.6__py3-none-any.whl → 0.1.17__py3-none-any.whl

agent/providers/omni/tool_manager.py

@@ -1,91 +0,0 @@
-# """Omni tool manager implementation."""
-
-# from typing import Dict, List, Type, Any
-
-# from computer import Computer
-# from ...core.tools import BaseToolManager, BashTool, EditTool
-
-# class OmniToolManager(BaseToolManager):
-#     """Tool manager for multi-provider support."""
-
-#     def __init__(self, computer: Computer):
-#         """Initialize Omni tool manager.
-
-#         Args:
-#             computer: Computer instance for tools
-#         """
-#         super().__init__(computer)
-
-#     def get_anthropic_tools(self) -> List[Dict[str, Any]]:
-#         """Get tools formatted for Anthropic API.
-
-#         Returns:
-#             List of tool parameters in Anthropic format
-#         """
-#         tools: List[Dict[str, Any]] = []
-
-#         # Map base tools to Anthropic format
-#         for tool in self.tools.values():
-#             if isinstance(tool, BashTool):
-#                 tools.append({
-#                     "type": "bash_20241022",
-#                     "name": tool.name
-#                 })
-#             elif isinstance(tool, EditTool):
-#                 tools.append({
-#                     "type": "text_editor_20241022",
-#                     "name": "str_replace_editor"
-#                 })
-
-#         return tools
-
-#     def get_openai_tools(self) -> List[Dict]:
-#         """Get tools formatted for OpenAI API.
-
-#         Returns:
-#             List of tool parameters in OpenAI format
-#         """
-#         tools = []
-
-#         # Map base tools to OpenAI format
-#         for tool in self.tools.values():
-#             tools.append({
-#                 "type": "function",
-#                 "function": tool.get_schema()
-#             })
-
-#         return tools
-
-#     def get_groq_tools(self) -> List[Dict]:
-#         """Get tools formatted for Groq API.
-
-#         Returns:
-#             List of tool parameters in Groq format
-#         """
-#         tools = []
-
-#         # Map base tools to Groq format
-#         for tool in self.tools.values():
-#             tools.append({
-#                 "type": "function",
-#                 "function": tool.get_schema()
-#             })
-
-#         return tools
-
-#     def get_qwen_tools(self) -> List[Dict]:
-#         """Get tools formatted for Qwen API.
-
-#         Returns:
-#             List of tool parameters in Qwen format
-#         """
-#         tools = []
-
-#         # Map base tools to Qwen format
-#         for tool in self.tools.values():
-#             tools.append({
-#                 "type": "function",
-#                 "function": tool.get_schema()
-#             })
-
-#         return tools

agent/providers/omni/visualization.py

@@ -1,130 +0,0 @@
-"""Visualization utilities for the Cua provider."""
-
-import base64
-import logging
-from io import BytesIO
-from typing import Tuple
-from PIL import Image, ImageDraw
-
-logger = logging.getLogger(__name__)
-
-
-def visualize_click(x: int, y: int, img_base64: str) -> Image.Image:
-    """Visualize a click action by drawing on the screenshot.
-
-    Args:
-        x: X coordinate of the click
-        y: Y coordinate of the click
-        img_base64: Base64 encoded image to draw on
-
-    Returns:
-        PIL Image with visualization
-    """
-    try:
-        # Decode the base64 image
-        img_data = base64.b64decode(img_base64)
-        img = Image.open(BytesIO(img_data))
-
-        # Create a drawing context
-        draw = ImageDraw.Draw(img)
-
-        # Draw concentric circles at the click position
-        small_radius = 10
-        large_radius = 30
-
-        # Draw filled inner circle
-        draw.ellipse(
-            [(x - small_radius, y - small_radius), (x + small_radius, y + small_radius)],
-            fill="red",
-        )
-
-        # Draw outlined outer circle
-        draw.ellipse(
-            [(x - large_radius, y - large_radius), (x + large_radius, y + large_radius)],
-            outline="red",
-            width=3,
-        )
-
-        return img
-
-    except Exception as e:
-        logger.error(f"Error visualizing click: {str(e)}")
-        # Return a blank image in case of error
-        return Image.new("RGB", (800, 600), color="white")
-
-
-def visualize_scroll(direction: str, clicks: int, img_base64: str) -> Image.Image:
-    """Visualize a scroll action by drawing arrows on the screenshot.
-
-    Args:
-        direction: 'up' or 'down'
-        clicks: Number of scroll clicks
-        img_base64: Base64 encoded image to draw on
-
-    Returns:
-        PIL Image with visualization
-    """
-    try:
-        # Decode the base64 image
-        img_data = base64.b64decode(img_base64)
-        img = Image.open(BytesIO(img_data))
-
-        # Get image dimensions
-        width, height = img.size
-
-        # Create a drawing context
-        draw = ImageDraw.Draw(img)
-
-        # Determine arrow direction and positions
-        center_x = width // 2
-        arrow_width = 100
-
-        if direction.lower() == "up":
-            # Draw up arrow in the middle of the screen
-            arrow_y = height // 2
-            # Arrow points
-            points = [
-                (center_x, arrow_y - 50),  # Top point
-                (center_x - arrow_width // 2, arrow_y + 50),  # Bottom left
-                (center_x + arrow_width // 2, arrow_y + 50),  # Bottom right
-            ]
-            color = "blue"
-        else:  # down
-            # Draw down arrow in the middle of the screen
-            arrow_y = height // 2
-            # Arrow points
-            points = [
-                (center_x, arrow_y + 50),  # Bottom point
-                (center_x - arrow_width // 2, arrow_y - 50),  # Top left
-                (center_x + arrow_width // 2, arrow_y - 50),  # Top right
-            ]
-            color = "green"
-
-        # Draw filled arrow
-        draw.polygon(points, fill=color)
-
-        # Add text showing number of clicks
-        text_y = arrow_y + 70 if direction.lower() == "down" else arrow_y - 70
-        draw.text((center_x - 40, text_y), f"{clicks} clicks", fill="black")
-
-        return img
-
-    except Exception as e:
-        logger.error(f"Error visualizing scroll: {str(e)}")
-        # Return a blank image in case of error
-        return Image.new("RGB", (800, 600), color="white")
-
-
-def calculate_element_center(box: Tuple[int, int, int, int]) -> Tuple[int, int]:
-    """Calculate the center coordinates of a bounding box.
-
-    Args:
-        box: Tuple of (left, top, right, bottom) coordinates
-
-    Returns:
-        Tuple of (center_x, center_y) coordinates
-    """
-    left, top, right, bottom = box
-    center_x = (left + right) // 2
-    center_y = (top + bottom) // 2
-    return center_x, center_y

agent/types/__init__.py

@@ -1,23 +0,0 @@
-"""Type definitions for the agent package."""
-
-from .base import HostConfig, TaskResult, Annotation
-from .messages import Message, Request, Response, StepMessage, DisengageMessage
-from .tools import ToolInvocation, ToolInvocationState, ClientAttachment, ToolResult
-
-__all__ = [
-    # Base types
-    "HostConfig",
-    "TaskResult",
-    "Annotation",
-    # Message types
-    "Message",
-    "Request",
-    "Response",
-    "StepMessage",
-    "DisengageMessage",
-    # Tool types
-    "ToolInvocation",
-    "ToolInvocationState",
-    "ClientAttachment",
-    "ToolResult",
-]

agent/types/base.py

@@ -1,41 +0,0 @@
-"""Base type definitions."""
-
-from enum import Enum, auto
-from typing import Dict, Any
-from pydantic import BaseModel, ConfigDict
-
-
-class HostConfig(BaseModel):
-    """Host configuration."""
-
-    model_config = ConfigDict(extra="forbid")
-    hostname: str
-    port: int
-
-    @property
-    def address(self) -> str:
-        return f"{self.hostname}:{self.port}"
-
-
-class TaskResult(BaseModel):
-    """Result of a task execution."""
-
-    model_config = ConfigDict(extra="forbid")
-    result: str
-    vnc_password: str
-
-
-class Annotation(BaseModel):
-    """Annotation metadata."""
-
-    model_config = ConfigDict(extra="forbid")
-    id: str
-    vm_url: str
-
-
-class AgentLoop(Enum):
-    """Enumeration of available loop types."""
-
-    ANTHROPIC = auto()  # Anthropic implementation
-    OMNI = auto()  # OmniLoop implementation
-    # Add more loop types as needed

agent/types/messages.py

@@ -1,36 +0,0 @@
-"""Message-related type definitions."""
-
-from typing import List, Dict, Any, Optional
-from pydantic import BaseModel, ConfigDict
-
-from .tools import ToolInvocation
-
-class Message(BaseModel):
-    """Base message type."""
-    model_config = ConfigDict(extra='forbid')
-    role: str
-    content: str
-    annotations: Optional[List[Dict[str, Any]]] = None
-    toolInvocations: Optional[List[ToolInvocation]] = None
-    data: Optional[List[Dict[str, Any]]] = None
-    errors: Optional[List[str]] = None
-
-class Request(BaseModel):
-    """Request type."""
-    model_config = ConfigDict(extra='forbid')
-    messages: List[Message]
-    selectedModel: str
-
-class Response(BaseModel):
-    """Response type."""
-    model_config = ConfigDict(extra='forbid')
-    messages: List[Message]
-    vm_url: str
-
-class StepMessage(Message):
-    """Message for a single step."""
-    pass
-
-class DisengageMessage(BaseModel):
-    """Message indicating disengagement."""
-    pass

cua_agent-0.1.6.dist-info/RECORD

@@ -1,64 +0,0 @@
-agent/README.md,sha256=8EFnLrKejthEcL9bZflQSbvA-KwpiPanBz8TEEwRub8,2153
-agent/__init__.py,sha256=Pil-INEbTU7iOlZXP3QN-kV_IHtW9uk7PuJCON319Ws,1437
-agent/core/README.md,sha256=VOXNVbR0ugxf9gCXYmZtUU2kngZhfi29haT_oSxK0Lk,3559
-agent/core/__init__.py,sha256=bds3kSkCILroIxxqU4scCPPCr-pooIkF2S4oU3OgsGY,638
-agent/core/callbacks.py,sha256=VbGIf5QkHh3Q0KsLM6wv7hRdIA5WExTVYLm64bckyUA,4306
-agent/core/computer_agent.py,sha256=7JOwAEeB1CL8Sw_1WAE3_a9tswZyCaBdXWDCfD3r7M4,9976
-agent/core/experiment.py,sha256=Ywj6q3JZFDKicfPuQsDl0vSN55HS7-Cnk3u3EcUCKe8,8866
-agent/core/loop.py,sha256=j4zI7h6mifQ5kMn9y0NcjAi1ZpvSBHoPL7Lk2e1OGsQ,9255
-agent/core/messages.py,sha256=N8pV8Eh-AJpMuDPRI5OGWUIOU6DRr-pQjK9XU0go9Hk,7637
-agent/core/telemetry.py,sha256=HElPd32k_w2SJ6t-Cc3j_2-AKdLbFwh2YlM8QViDgRw,4790
-agent/core/tools/__init__.py,sha256=xZen-PqUp2dUaMEHJowXCQm33_5Sxhsx9PSoD0rq6tI,489
-agent/core/tools/base.py,sha256=CdzRFNuOjNfzgyTUN4ZoCGkUDR5HI0ECQVpvrUdEij8,2295
-agent/core/tools/bash.py,sha256=jnJKVlHn8np8e0gWd8EO0_qqjMkfQzutSugA_Iol4jE,1585
-agent/core/tools/collection.py,sha256=NuwTn6dXSyznxWodfmFDQwUlxxaGb4oBPym4AEJABSQ,1338
-agent/core/tools/computer.py,sha256=lT_aW3huoYpcM8kffuokELupSz_WZG_qkaW1gITRC58,3892
-agent/core/tools/edit.py,sha256=kv4jTKCM0VXrnoNErf7mT-xlr81-7T8v49_VA9y_L4Y,2005
-agent/core/tools/manager.py,sha256=IRsCXjGc076nncQuyIjODoafnHTDhrf9sP5B4q5Pcdo,1742
-agent/providers/__init__.py,sha256=b4tIBAaIB1V7p8V0BWipHVnMhfHH_OuVgP4OWGSHdD8,194
-agent/providers/anthropic/__init__.py,sha256=Mj11IZnVshZ2iHkvg4Z5-jrQIaD1WvzDz2Zk_pMwqIA,149
-agent/providers/anthropic/api/client.py,sha256=Y_g4Xg8Ko4tCqjipVm0GBMw-86vw0KQVXS5aWzJinzw,7038
-agent/providers/anthropic/api/logging.py,sha256=vHpwkIyOZdkSTVIH4ycbBPd4a_rzhP7Osu1I-Ayouwc,5154
-agent/providers/anthropic/callbacks/manager.py,sha256=dRKN7MuBze2dLal0iHDxCKYqMdh_KShSphuwn7zC-c4,1878
-agent/providers/anthropic/loop.py,sha256=uPjgXoGRdJb5bsJchUh_0aUuyRBm-HSp7jaM2cKg61I,19466
-agent/providers/anthropic/messages/manager.py,sha256=6FobzAHh5-7dxaxbUdG1--1UY4w-mh3MFytX6ONrK3c,4972
-agent/providers/anthropic/prompts.py,sha256=nHFfgPrfvnWrEdVP7EUBGUHAI85D2X9HeZirk9EwncU,1941
-agent/providers/anthropic/tools/__init__.py,sha256=JyZwuVtPUnZwRSZBSCdQv9yxbLCsygm3l8Ywjjt9qTQ,661
-agent/providers/anthropic/tools/base.py,sha256=WnRDbqO25tQzLpS2RU2ZXTLF5wd5IqU7SiyRAglQat4,2752
-agent/providers/anthropic/tools/bash.py,sha256=CIh4pO0jEdSZApnjpmFhrQbTTiwxivuOgv1-QLN0Ydw,5740
-agent/providers/anthropic/tools/collection.py,sha256=RBK_6hxfHExR-EOxadiLl0OznmFj07nyIUjFgaYZ6Eo,960
-agent/providers/anthropic/tools/computer.py,sha256=vYni1jDOOgzSSBOJxHcEKxvKUYRp5_nQ-9dmpGdLwm4,25858
-agent/providers/anthropic/tools/edit.py,sha256=EGRP61MDA4Oue1D7Q-_vLpd6LdGbdBA1Z4HSZ66DbmI,13465
-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/omni/__init__.py,sha256=eTUh4Pmh4zO-RLnP-wAFm8EkJBMImT-G2xnVIYWRti0,744
-agent/providers/omni/callbacks.py,sha256=ZG9NCgsHWt6y5jKsfcGLaoLxTpmKnIhCArDdeP4q9sA,2369
-agent/providers/omni/clients/anthropic.py,sha256=X_QRVxqwA_ExdUqgBEwo1aHOfZQxVIBDmDugNHF97OM,3554
-agent/providers/omni/clients/base.py,sha256=zAAgPi0jl3SWPC730R9l79E8bfYPSo39UtCSE-mrK6I,1076
-agent/providers/omni/clients/groq.py,sha256=HEinpE0_Cp_-geMyjJ8qaTPl0regPtETPkem4U13qG4,3599
-agent/providers/omni/clients/openai.py,sha256=E4TAXMUFoYTunJETCWCNx5XAc6xutiN4rB6PlVpzC5s,5972
-agent/providers/omni/clients/utils.py,sha256=Ani9CVVBm_J2Dl51WG6p1GVuoI6cq8scISrG0pmQ37o,688
-agent/providers/omni/experiment.py,sha256=ZZ45U5NEkpzMNeMO9hJfpRf3iPNFaSZVwChcfqOgbI0,10002
-agent/providers/omni/image_utils.py,sha256=qIFuNi5cIMVwrqYBXG1T6PxUlbxz7gIngFFP39bZIlU,2782
-agent/providers/omni/loop.py,sha256=10GxyZFG8wAYObaaInWSZDRVwWNnZk_qhqdGr3PIPe0,44022
-agent/providers/omni/messages.py,sha256=zdjQCAMH-hOyrQQesHhTiIsQbw43KqVSmVIzS8JOIFA,6134
-agent/providers/omni/parser.py,sha256=4n1rzaD-mHi7sMfeqChgOyrJuciwzL95x32BGI6GATM,9194
-agent/providers/omni/prompts.py,sha256=Mupjy0bUwBjcAeLXpE1r1jisYPSlhwsp-IXJKEKrEtw,3779
-agent/providers/omni/tool_manager.py,sha256=O6DxyEI-Vg6jt99phh011o4q4me_vNhH2YffIxkO4GM,2585
-agent/providers/omni/tools/__init__.py,sha256=RkxsPTow3jpOKuXJ1ZKb-KBi6lbxGWfjC9gaV6hSZIs,278
-agent/providers/omni/tools/bash.py,sha256=y_ibfP9iRcbiU_E0faAoa4DCP_BlkMlKOOURdBBIGZE,2030
-agent/providers/omni/tools/computer.py,sha256=s8WVA_xGROEfdmCYjEqr563ySp4DRMlsLVuu54nH0Ww,9129
-agent/providers/omni/tools/manager.py,sha256=EyNABQeJc8bEcJ4hFsRodKyBNkZpThfOHk1Ku3Pzsg8,2519
-agent/providers/omni/types.py,sha256=rpr7-mH9VK1R-nJ6tVu1gKp427j-hw1DpHc197b44nU,1017
-agent/providers/omni/utils.py,sha256=X2IBki6yJQFaEz7PDjkx8CqQq2R1v7nldRcOT5j7YcA,6381
-agent/providers/omni/visualization.py,sha256=N3qVQLxYmia3iSVC5oCt5YRlMPuVfylCOyB99R33u8U,3924
-agent/telemetry.py,sha256=pVGxbj0ewnvq4EGj28CydN4a1iOfvZR_XKL3vIOqhOM,390
-agent/types/__init__.py,sha256=Xi6nNSsfbsGxs3We8gbdEY0ew0Jf0A0Prs5393Tvveg,568
-agent/types/base.py,sha256=_5LNleRTqoL55VHEEZntL8x-OQom2A3FiTf2ttdM_HQ,857
-agent/types/messages.py,sha256=4-hwtxeAhto90_EZpHFducddtsHUsHauvXzYrpKG4RE,953
-agent/types/tools.py,sha256=Jes2CFCFqC727WWHbO-sG7V03rBHnQe5X7Oi9ZkuScI,877
-cua_agent-0.1.6.dist-info/METADATA,sha256=rD_j8q7aC5wkNQtpbgRLyo3-5z_zCDivJE8MwyPpz6I,4528
-cua_agent-0.1.6.dist-info/WHEEL,sha256=thaaA2w1JzcGC48WYufAs8nrYZjJm8LqNfnXFOFyCC4,90
-cua_agent-0.1.6.dist-info/entry_points.txt,sha256=6OYgBcLyFCUgeqLgnvMyOJxPCWzgy7se4rLPKtNonMs,34
-cua_agent-0.1.6.dist-info/RECORD,,