camel-ai 0.2.73a12__tar.gz → 0.2.74__tar.gz

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: camel-ai
-Version: 0.2.73a12
+Version: 0.2.74
 Summary: Communicative Agents for AI Society Study
 Project-URL: Homepage, https://www.camel-ai.org/
 Project-URL: Repository, https://github.com/camel-ai/camel
@@ -237,7 +237,7 @@ Requires-Dist: exa-py<2,>=1.10.0; extra == 'eigent'
 Requires-Dist: ffmpeg-python<0.3,>=0.2.0; extra == 'eigent'
 Requires-Dist: html2text>=2024.2.26; extra == 'eigent'
 Requires-Dist: imageio[pyav]<3,>=2.34.2; extra == 'eigent'
-Requires-Dist: markitdown>=0.1.1; extra == 'eigent'
+Requires-Dist: markitdown[all]>=0.1.1; extra == 'eigent'
 Requires-Dist: mcp-server-fetch==2025.1.17; extra == 'eigent'
 Requires-Dist: mcp-simple-arxiv==0.2.2; extra == 'eigent'
 Requires-Dist: numpy<=2.2,>=1.2; extra == 'eigent'
@@ -254,6 +254,7 @@ Requires-Dist: python-pptx>=1.0.2; extra == 'eigent'
 Requires-Dist: reportlab>=4.4.2; extra == 'eigent'
 Requires-Dist: requests-oauthlib<2,>=1.3.1; extra == 'eigent'
 Requires-Dist: scenedetect>=0.6.5.2; extra == 'eigent'
+Requires-Dist: slack-sdk<4,>=3.27.2; extra == 'eigent'
 Requires-Dist: tabulate>=0.9.0; extra == 'eigent'
 Requires-Dist: websockets<15.1,>=13.0; extra == 'eigent'
 Requires-Dist: wikipedia<2,>=1; extra == 'eigent'

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/camel/__init__.py

@@ -14,7 +14,7 @@
 
 from camel.logger import disable_logging, enable_logging, set_log_level
 
-__version__ = '0.2.73a12'
+__version__ = '0.2.74'
 
 __all__ = [
     '__version__',

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/camel/interpreters/e2b_interpreter.py

@@ -28,6 +28,11 @@ class E2BInterpreter(BaseInterpreter):
     Args:
         require_confirm (bool, optional): If True, prompt user before running
             code strings for security. (default: :obj:`True`)
+
+    Environment Variables:
+        E2B_API_KEY: The API key for authenticating with the E2B service.
+        E2B_DOMAIN: The base URL for the E2B API. If not provided,
+            will use the default E2B endpoint.
     """
 
     _CODE_TYPE_MAPPING: ClassVar[Dict[str, Optional[str]]] = {
@@ -55,7 +60,35 @@ class E2BInterpreter(BaseInterpreter):
         from e2b_code_interpreter import Sandbox
 
         self.require_confirm = require_confirm
-        self._sandbox = Sandbox(api_key=os.environ.get("E2B_API_KEY"))
+
+        # Get API key from environment variable
+        api_key = os.environ.get("E2B_API_KEY")
+
+        # Get domain from environment variable
+        domain = os.environ.get("E2B_DOMAIN")
+
+        # Create sandbox with appropriate parameters
+        sandbox_kwargs = {"api_key": api_key}
+
+        # Only add domain if it's provided
+        # (to maintain compatibility with standard E2B)
+        if domain:
+            sandbox_kwargs["domain"] = domain
+            logger.info(f"Using custom E2B endpoint: {domain}")
+
+        try:
+            self._sandbox = Sandbox(**sandbox_kwargs)
+        except TypeError as e:
+            if domain and "domain" in str(e):
+                logger.warning(
+                    f"The e2b_code_interpreter library doesn't support "
+                    f"custom domain. "
+                    f"Using default E2B endpoint. Error: {e}"
+                )
+                # Fallback to default configuration without domain
+                self._sandbox = Sandbox(api_key=api_key)
+            else:
+                raise e
 
     def __del__(self) -> None:
         r"""Destructor for the E2BInterpreter class.

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/camel/models/anthropic_model.py

@@ -21,8 +21,8 @@ from camel.messages import OpenAIMessage
 from camel.models.openai_compatible_model import OpenAICompatibleModel
 from camel.types import ChatCompletion, ChatCompletionChunk, ModelType
 from camel.utils import (
-    AnthropicTokenCounter,
     BaseTokenCounter,
+    OpenAITokenCounter,
     api_keys_required,
     dependencies_required,
 )
@@ -141,11 +141,13 @@ class AnthropicModel(OpenAICompatibleModel):
         r"""Initialize the token counter for the model backend.
 
         Returns:
-            BaseTokenCounter: The token counter following the model's
+            OpenAITokenCounter: The token counter following the model's
                 tokenization style.
         """
+        # TODO: use anthropic token counter
+
         if not self._token_counter:
-            self._token_counter = AnthropicTokenCounter(self.model_type)
+            self._token_counter = OpenAITokenCounter(ModelType.GPT_4O_MINI)
         return self._token_counter
 
     def check_model_config(self):

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/camel/societies/workforce/prompts.py

@@ -198,10 +198,9 @@ Now you should summarize the scenario and return the result of the task.
 
 TASK_DECOMPOSE_PROMPT = r"""You need to either decompose a complex task or enhance a simple one, following these important principles to maximize efficiency and clarity for the executing agents:
 
-0.  **Enhance Simple Tasks, Don't Decompose**: First, evaluate if the task is a single, straightforward action.
-    *   **If the task is simple, DO NOT decompose it.** Instead, **rewrite and enhance** it by consolidating the main goal with any specific details from the additional information into a single, self-contained instruction. The goal is to produce a high-quality task with a clear, specific deliverable.
-    *   For example, a vague task like "Generate a report about camel-ai" should be enhanced to "Generate a comprehensive, well-structured report about camel-ai as one HTML file".
-    *   **Only decompose if a task is truly complex**: A task is complex if it involves multiple distinct steps, requires different worker skills, or can be significantly sped up by running parts in parallel.
+0.  **Analyze Task Complexity**: First, evaluate if the task is a single, straightforward action or a complex one.
+    *   **If the task is complex or could be decomposed into multiple subtasks run in parallel, decompose it.** A task is considered complex if it involves multiple distinct steps, requires different skills, or can be significantly sped up by running parts in parallel.
+    *   **If the task is simple, do not decompose it.** Instead, **rewrite and enhance** it to produce a high-quality task with a clear, specific deliverable.
 
 1.  **Self-Contained Subtasks** (if decomposing): This is critical principle. Each subtask's description **must be fully self-sufficient and independently understandable**. The agent executing the subtask has **no knowledge** of the parent task, other subtasks, or the overall workflow.
     *   **DO NOT** use relative references like "the first task," "the paper mentioned above," or "the result from the previous step."
@@ -230,21 +229,6 @@ These principles aim to reduce overall completion time by maximizing concurrent
 
 **EXAMPLE FORMAT ONLY** (DO NOT use this example content for actual task decomposition):
 
-***
-**Example 0: Simple Task (Enhance, Don't Decompose)**
-
-*   **Overall Task**: "Generate a report about camel-ai"
-*   **Available Workers**:
-    *   `DocumentAgent`: A worker that can read documents and write summaries.
-*   **Correct Output**:
-    ```xml
-    <tasks>
-    <task>Generate a comprehensive, well-structured report about camel-ai as one HTML file</task>
-    </tasks>
-    ```
-*   **Reasoning**: The original task is simple but its details are vague. It is **enhanced** by consolidating all given details (output format, content focus) into a single, precise, and self-contained instruction. This follows the "Enhance Simple Tasks" principle and does not invent new information.
-***
-
 ***
 **Example 1: Sequential Task for a Single Worker**
 

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/camel/societies/workforce/workforce.py

@@ -84,7 +84,7 @@ logger = get_logger(__name__)
 # Constants for configuration values
 MAX_TASK_RETRIES = 3
 MAX_PENDING_TASKS_LIMIT = 20
-TASK_TIMEOUT_SECONDS = 180.0
+TASK_TIMEOUT_SECONDS = 480.0
 DEFAULT_WORKER_POOL_SIZE = 10
 
 
@@ -163,6 +163,11 @@ class Workforce(BaseNode):
             for graceful shutdown when a task fails 3 times. During this
             period, the workforce remains active for debugging.
             Set to 0 for immediate shutdown. (default: :obj:`15.0`)
+        task_timeout_seconds (Optional[float], optional): The timeout in
+            seconds for waiting for tasks to be returned by workers. If None,
+            uses the global TASK_TIMEOUT_SECONDS value (600.0 seconds).
+            Increase this value for tasks that require more processing time.
+            (default: :obj:`None`)
         share_memory (bool, optional): Whether to enable shared memory across
             SingleAgentWorker instances in the workforce. When enabled, all
             SingleAgentWorker instances, coordinator agent, and task planning
@@ -231,6 +236,7 @@ class Workforce(BaseNode):
         graceful_shutdown_timeout: float = 15.0,
         share_memory: bool = False,
         use_structured_output_handler: bool = True,
+        task_timeout_seconds: Optional[float] = None,
     ) -> None:
         super().__init__(description)
         self._child_listening_tasks: Deque[
@@ -241,6 +247,9 @@ class Workforce(BaseNode):
         self.graceful_shutdown_timeout = graceful_shutdown_timeout
         self.share_memory = share_memory
         self.use_structured_output_handler = use_structured_output_handler
+        self.task_timeout_seconds = (
+            task_timeout_seconds or TASK_TIMEOUT_SECONDS
+        )
         if self.use_structured_output_handler:
             self.structured_handler = StructuredOutputHandler()
         self.metrics_logger = WorkforceLogger(workforce_id=self.node_id)
@@ -343,12 +352,7 @@ class Workforce(BaseNode):
         # Set up task agent with default system message and required tools
         task_sys_msg = BaseMessage.make_assistant_message(
             role_name="Task Planner",
-            content="You are going to compose and decompose tasks. Keep "
-            "tasks that are sequential and require the same type of "
-            "agent together in one agent process. Only decompose tasks "
-            "that can be handled in parallel and require different types "
-            "of agents. This ensures efficient execution by minimizing "
-            "context switching between agents.",
+            content="You are going to handle tasks.",
         )
         task_planning_tools = TaskPlanningToolkit().get_tools()
 
@@ -2311,7 +2315,7 @@ class Workforce(BaseNode):
             # Add timeout to prevent indefinite waiting
             return await asyncio.wait_for(
                 self._channel.get_returned_task_by_publisher(self.node_id),
-                timeout=TASK_TIMEOUT_SECONDS,
+                timeout=self.task_timeout_seconds,
             )
         except Exception as e:
             error_msg = (
@@ -3066,6 +3070,7 @@ class Workforce(BaseNode):
             graceful_shutdown_timeout=self.graceful_shutdown_timeout,
             share_memory=self.share_memory,
             use_structured_output_handler=self.use_structured_output_handler,
+            task_timeout_seconds=self.task_timeout_seconds,
         )
 
         for child in self._children:

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/camel/toolkits/hybrid_browser_toolkit/config_loader.py

@@ -23,6 +23,7 @@ class BrowserConfig:
     headless: bool = True
     user_data_dir: Optional[str] = None
     stealth: bool = False
+    console_log_limit: int = 1000
 
     # Default settings
     default_start_url: str = "https://google.com/"
@@ -105,6 +106,8 @@ class ConfigLoader:
                 browser_kwargs["connect_over_cdp"] = value
             elif key == "cdpUrl":
                 browser_kwargs["cdp_url"] = value
+            elif key == "consoleLogLimit":
+                browser_kwargs["console_log_limit"] = value
             elif key == "cacheDir":
                 toolkit_kwargs["cache_dir"] = value
             elif key == "browserLogToFile":

{camel_ai-0.2.73a12 → camel_ai-0.2.74}/camel/toolkits/hybrid_browser_toolkit/hybrid_browser_toolkit_ts.py

@@ -64,11 +64,16 @@ class HybridBrowserToolkit(BaseToolkit, RegisteredAgentToolkit):
         "browser_select",
         "browser_scroll",
         "browser_enter",
+        "browser_mouse_control",
+        "browser_mouse_drag",
+        "browser_press_key",
         "browser_wait_user",
         "browser_solve_task",
         "browser_switch_tab",
         "browser_close_tab",
         "browser_get_tab_info",
+        "browser_console_view",
+        "browser_console_exec",
     ]
 
     def __init__(
@@ -863,6 +868,156 @@ class HybridBrowserToolkit(BaseToolkit, RegisteredAgentToolkit):
                 "total_tabs": 0,
             }
 
+    async def browser_mouse_control(
+        self, *, control: str, x: float, y: float
+    ) -> Dict[str, Any]:
+        r"""Control the mouse to interact with browser with x, y coordinates
+
+        Args:
+            control ([str]): The action to perform: 'click', 'right_click'
+            or 'dblclick'.
+            x (float): x-coordinate for the control action.
+            y (float): y-coordinate for the control action.
+
+        Returns:
+            Dict[str, Any]: A dictionary with the result of the action:
+                - "result" (str): Confirmation of the action.
+                - "snapshot" (str): A snapshot of the page after mouse
+                control action.
+                - "tabs" (List[Dict]): Information about all open tabs.
+                - "current_tab" (int): Index of the active tab.
+                - "total_tabs" (int): Total number of open tabs.
+        """
+        try:
+            ws_wrapper = await self._get_ws_wrapper()
+            result = await ws_wrapper.mouse_control(control, x, y)
+
+            # Add tab information
+            tab_info = await ws_wrapper.get_tab_info()
+            result.update(
+                {
+                    "tabs": tab_info,
+                    "current_tab": next(
+                        (
+                            i
+                            for i, tab in enumerate(tab_info)
+                            if tab.get("is_current")
+                        ),
+                        0,
+                    ),
+                    "total_tabs": len(tab_info),
+                }
+            )
+
+            return result
+        except Exception as e:
+            logger.error(f"Failed to control mouse: {e}")
+            return {
+                "result": f"Error with mouse control: {e}",
+                "snapshot": "",
+                "tabs": [],
+                "current_tab": 0,
+                "total_tabs": 0,
+            }
+
+    async def browser_mouse_drag(
+        self, *, from_ref: str, to_ref: str
+    ) -> Dict[str, Any]:
+        r"""Control the mouse to drag and drop in the browser using ref IDs.
+
+        Args:
+            from_ref (str): The `ref` ID of the source element to drag from.
+            to_ref (str): The `ref` ID of the target element to drag to.
+
+        Returns:
+            Dict[str, Any]: A dictionary with the result of the action:
+                - "result" (str): Confirmation of the action.
+                - "snapshot" (str): A new page snapshot.
+                - "tabs" (List[Dict]): Information about all open tabs.
+                - "current_tab" (int): Index of the active tab.
+                - "total_tabs" (int): Total number of open tabs.
+        """
+        try:
+            ws_wrapper = await self._get_ws_wrapper()
+            result = await ws_wrapper.mouse_drag(from_ref, to_ref)
+
+            # Add tab information
+            tab_info = await ws_wrapper.get_tab_info()
+            result.update(
+                {
+                    "tabs": tab_info,
+                    "current_tab": next(
+                        (
+                            i
+                            for i, tab in enumerate(tab_info)
+                            if tab.get("is_current")
+                        ),
+                        0,
+                    ),
+                    "total_tabs": len(tab_info),
+                }
+            )
+
+            return result
+        except Exception as e:
+            logger.error(f"Error with mouse drag and drop: {e}")
+            return {
+                "result": f"Error with mouse drag and drop: {e}",
+                "snapshot": "",
+                "tabs": [],
+                "current_tab": 0,
+                "total_tabs": 0,
+            }
+
+    async def browser_press_key(self, *, keys: List[str]) -> Dict[str, Any]:
+        r"""Press key and key combinations.
+        Supports single key press or combination of keys by concatenating
+        them with '+' separator.
+
+        Args:
+            keys (List[str]): key or list of keys.
+
+        Returns:
+            Dict[str, Any]: A dictionary with the result of the action:
+                - "result" (str): Confirmation of the action.
+                - "snapshot" (str): A snapshot of the page after
+                press key action.
+                - "tabs" (List[Dict]): Information about all open tabs.
+                - "current_tab" (int): Index of the active tab.
+                - "total_tabs" (int): Total number of open tabs.
+        """
+        try:
+            ws_wrapper = await self._get_ws_wrapper()
+            result = await ws_wrapper.press_key(keys)
+
+            # Add tab information
+            tab_info = await ws_wrapper.get_tab_info()
+            result.update(
+                {
+                    "tabs": tab_info,
+                    "current_tab": next(
+                        (
+                            i
+                            for i, tab in enumerate(tab_info)
+                            if tab.get("is_current")
+                        ),
+                        0,
+                    ),
+                    "total_tabs": len(tab_info),
+                }
+            )
+
+            return result
+        except Exception as e:
+            logger.error(f"Failed to press key: {e}")
+            return {
+                "result": f"Error with press key: {e}",
+                "snapshot": "",
+                "tabs": [],
+                "current_tab": 0,
+                "total_tabs": 0,
+            }
+
     async def browser_switch_tab(self, *, tab_id: str) -> Dict[str, Any]:
         r"""Switches to a different browser tab using its ID.
 
@@ -1002,6 +1157,71 @@ class HybridBrowserToolkit(BaseToolkit, RegisteredAgentToolkit):
                 "total_tabs": 0,
             }
 
+    async def browser_console_view(self) -> Dict[str, Any]:
+        r"""View current page console logs.
+
+        Returns:
+            Dict[str, Any]: A dictionary with tab information:
+                - "console_messages" (List[Dict]) : List of messages logged
+                in the current page
+
+        """
+        try:
+            ws_wrapper = await self._get_ws_wrapper()
+            console_logs = await ws_wrapper.console_view()
+
+            return {"console_messages": console_logs}
+        except Exception as e:
+            logger.error(f"Failed to get console view: {e}")
+            return {"console_messages": []}
+
+    async def browser_console_exec(self, code: str) -> Dict[str, Any]:
+        r"""Execute javascript code in the console of the current page and get
+        results.
+
+        Args:
+            code (str): JavaScript code to execute in the browser console.
+
+        Returns:
+            Dict[str, Any]: A dictionary with the result of the action:
+                - "result" (str): Confirmation of the action.
+                - "snapshot" (str): A snapshot of the active tab after
+                console execute action.
+                - "tabs" (List[Dict]): Information about remaining tabs.
+                - "current_tab" (int): Index of the new active tab.
+                - "total_tabs" (int): Total number of remaining tabs.
+        """
+        try:
+            ws_wrapper = await self._get_ws_wrapper()
+            result = await ws_wrapper.console_exec(code)
+
+            tab_info = await ws_wrapper.get_tab_info()
+            result.update(
+                {
+                    "tabs": tab_info,
+                    "current_tab": next(
+                        (
+                            i
+                            for i, tab in enumerate(tab_info)
+                            if tab.get("is_current")
+                        ),
+                        0,
+                    ),
+                    "total_tabs": len(tab_info),
+                }
+            )
+
+            return result
+        except Exception as e:
+            logger.error(f"Failed to execute javascript in console: {e}")
+            return {
+                "result": f"Error in code execution: {e}",
+                "snapshot": "",
+                "tabs": [],
+                "current_tab": 0,
+                "total_tabs": 0,
+            }
+
     # Additional methods for backward compatibility
     async def browser_wait_user(
         self, timeout_sec: Optional[float] = None
@@ -1146,10 +1366,15 @@ class HybridBrowserToolkit(BaseToolkit, RegisteredAgentToolkit):
             "browser_select": self.browser_select,
             "browser_scroll": self.browser_scroll,
             "browser_enter": self.browser_enter,
+            "browser_mouse_click": self.browser_mouse_control,
+            "browser_mouse_drag": self.browser_mouse_drag,
+            "browser_press_key": self.browser_press_key,
             "browser_wait_user": self.browser_wait_user,
             "browser_switch_tab": self.browser_switch_tab,
             "browser_close_tab": self.browser_close_tab,
             "browser_get_tab_info": self.browser_get_tab_info,
+            "browser_console_view": self.browser_console_view,
+            "browser_console_exec": self.browser_console_exec,
         }
 
         enabled_tools = []