lybic-guiagents 0.2.2__tar.gz → 0.2.3__tar.gz

{lybic_guiagents-0.2.2 → lybic_guiagents-0.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: lybic-guiagents
-Version: 0.2.2
+Version: 0.2.3
 Summary: An open-source agentic framework that enables AI to use computers like humans and can provide a multi-agent runtime environment as an infrastructure capability
 Author: Lybic Development Team
 Author-email: Lybic Development Team <lybic@tingyutech.com>
@@ -23,13 +23,21 @@ Requires-Dist: boto3
 Requires-Dist: scikit-learn
 Requires-Dist: tiktoken
 Requires-Dist: pydantic
-Requires-Dist: lybic>=0.7.3
+Requires-Dist: lybic<1.0.0,>=0.8.0b2
+Requires-Dist: grpcio
+Requires-Dist: grpcio-tools>=1.71.2
 Requires-Dist: pyobjc; platform_system == "Darwin"
 Requires-Dist: oa_atomacos; platform_system == "Darwin"
 Requires-Dist: pywinauto; platform_system == "Windows"
 Requires-Dist: pywin32; platform_system == "Windows"
+Requires-Dist: pytest>=8.4.2
+Requires-Dist: protobuf>=5.29.5
+Requires-Dist: googleapis-common-protos>=1.70.0
 Provides-Extra: dev
 Requires-Dist: pytest; extra == "dev"
+Provides-Extra: grpc
+Requires-Dist: grpcio; extra == "grpc"
+Requires-Dist: grpcio-tools>=1.71.2; extra == "grpc"
 Provides-Extra: vmware-and-evaluation
 Requires-Dist: python-pptx; extra == "vmware-and-evaluation"
 Requires-Dist: odfpy; extra == "vmware-and-evaluation"
@@ -148,6 +156,8 @@ Requires-Dist: pyobjc; platform_system == "Darwin" and extra == "all"
 Requires-Dist: oa_atomacos; platform_system == "Darwin" and extra == "all"
 Requires-Dist: pywinauto; platform_system == "Windows" and extra == "all"
 Requires-Dist: pywin32; platform_system == "Windows" and extra == "all"
+Requires-Dist: grpcio; extra == "all"
+Requires-Dist: grpcio-tools>=1.71.2; extra == "all"
 Dynamic: author
 Dynamic: license-file
 Dynamic: requires-python
@@ -372,6 +382,13 @@ python gui_agents/cli_app.py --backend pyautogui --mode fast
 > [!WARNING]
 > The agent will directly control your computer with `--backend pyautogui`. Please use with care.
 
+### Docker
+You can also run Lybic GUI Agent using Docker. This is an example of how to run it with the `lybic` backend:
+```sh
+docker run --rm -it --env-file gui_agents/.env agenticlybic/guiagent --backend lybic
+```
+> **Note**: This command starts the agent in interactive mode. The `--env-file` flag points to the environment file. Please ensure the path is correct.
+
 ### Lybic Sandbox Configuration
 
 The simplest way to configure Lybic Sandbox is still to edit the `.env` file and add your API keys, as mentioned in the [API Key Configuration](#api-key-configuration) section.
@@ -392,15 +409,19 @@ LYBIC_MAX_LIFE_SECONDS=3600
 
 ### Use as a service
 
-After installing lybic-guiagents, you can run it as a service.
+You can interact with the agent programmatically either by importing it as a Python library or by running it as a gRPC service.
+
+#### As a Python Library
+
+After installing `lybic-guiagents`, you can import and use its components directly in your Python code.
 
 Main Components:
-- AgentService: High-level service interface (recommended for most users)
-- AgentS2, AgentSFast: Core agent implementations  
-- HardwareInterface: Hardware abstraction layer
-- ServiceConfig: Configuration management
+- `AgentService`: High-level service interface (recommended for most users).
+- `AgentS2`, `AgentSFast`: Core agent implementations.
+- `HardwareInterface`: Hardware abstraction layer for controlling the GUI.
+- `ServiceConfig`: Configuration management.
 
-Quick Start:
+**Quick Start:**
 
 ```python
 from gui_agents import AgentService
@@ -409,6 +430,63 @@ result = service.execute_task("Take a screenshot")
 print(f"Task completed: {result.status}")
 ```
 
+#### As a gRPC Service
+
+You can also run the agent as a standalone gRPC service, which is ideal for distributed architectures or integrating with applications written in other languages.
+
+**1. Running the gRPC Server**
+
+First, run the gRPC server using Docker. This command overrides the default CLI entrypoint and starts the gRPC service on port 50051.
+
+```sh
+docker run --rm -it -p 50051:50051 --env-file gui_agents/.env agenticlybic/guiagent /app/.venv/bin/lybic-guiagent-grpc
+```
+> **Note**: The `-p 50051:50051` flag maps the container's gRPC port to your host machine.
+
+**2. Python Client Example**
+
+Once the service is running, you can interact with it using a gRPC client. Here is a Python example of how to send an instruction to the agent and stream its progress.
+
+First, ensure you have the necessary gRPC libraries and generated protobuf stubs:
+```sh
+# Install gRPC tools
+pip install grpcio grpcio-tools
+
+# Generate stubs from the .proto file
+python -m grpc_tools.protoc -Igui_agents/proto --python_out=gui_agents/proto/pb --grpc_python_out=gui_agents/proto/pb --pyi_out=gui_agents/proto/pb gui_agents/proto/agent.proto
+```
+
+Then, you can use the following script to communicate with the agent:
+
+```python
+import asyncio
+import grpc
+from gui_agents.proto.pb import agent_pb2, agent_pb2_grpc
+
+async def run_agent_instruction():
+    # Connect to the gRPC server
+    async with grpc.aio.insecure_channel('localhost:50051') as channel:
+        # Create a stub for the Agent service
+        stub = agent_pb2_grpc.AgentStub(channel)
+
+        # Create a request to run an instruction
+        request = agent_pb2.RunAgentInstructionRequest(
+            instruction="Open a calculator and compute 1 + 1"
+        )
+
+        print(f"Sending instruction: '{request.instruction}'")
+
+        # Call the RunAgentInstruction RPC and iterate over the stream of responses
+        try:
+            async for response in stub.RunAgentInstruction(request):
+                print(f"[{response.stage}] {response.message}")
+        except grpc.aio.AioRpcError as e:
+            print(f"An error occurred: {e.details()}")
+
+if __name__ == '__main__':
+    asyncio.run(run_agent_instruction())
+```
+
 ### VMware Configuration
 
 To use PyAutoGUI with VMware, you need to install [VMware Workstation Pro](https://www.vmware.com/products/desktop-hypervisor/workstation-and-fusion) (on Windows) and create a virtual machine. 

{lybic_guiagents-0.2.2 → lybic_guiagents-0.2.3}/README.md

@@ -218,6 +218,13 @@ python gui_agents/cli_app.py --backend pyautogui --mode fast
 > [!WARNING]
 > The agent will directly control your computer with `--backend pyautogui`. Please use with care.
 
+### Docker
+You can also run Lybic GUI Agent using Docker. This is an example of how to run it with the `lybic` backend:
+```sh
+docker run --rm -it --env-file gui_agents/.env agenticlybic/guiagent --backend lybic
+```
+> **Note**: This command starts the agent in interactive mode. The `--env-file` flag points to the environment file. Please ensure the path is correct.
+
 ### Lybic Sandbox Configuration
 
 The simplest way to configure Lybic Sandbox is still to edit the `.env` file and add your API keys, as mentioned in the [API Key Configuration](#api-key-configuration) section.
@@ -238,15 +245,19 @@ LYBIC_MAX_LIFE_SECONDS=3600
 
 ### Use as a service
 
-After installing lybic-guiagents, you can run it as a service.
+You can interact with the agent programmatically either by importing it as a Python library or by running it as a gRPC service.
+
+#### As a Python Library
+
+After installing `lybic-guiagents`, you can import and use its components directly in your Python code.
 
 Main Components:
-- AgentService: High-level service interface (recommended for most users)
-- AgentS2, AgentSFast: Core agent implementations  
-- HardwareInterface: Hardware abstraction layer
-- ServiceConfig: Configuration management
+- `AgentService`: High-level service interface (recommended for most users).
+- `AgentS2`, `AgentSFast`: Core agent implementations.
+- `HardwareInterface`: Hardware abstraction layer for controlling the GUI.
+- `ServiceConfig`: Configuration management.
 
-Quick Start:
+**Quick Start:**
 
 ```python
 from gui_agents import AgentService
@@ -255,6 +266,63 @@ result = service.execute_task("Take a screenshot")
 print(f"Task completed: {result.status}")
 ```
 
+#### As a gRPC Service
+
+You can also run the agent as a standalone gRPC service, which is ideal for distributed architectures or integrating with applications written in other languages.
+
+**1. Running the gRPC Server**
+
+First, run the gRPC server using Docker. This command overrides the default CLI entrypoint and starts the gRPC service on port 50051.
+
+```sh
+docker run --rm -it -p 50051:50051 --env-file gui_agents/.env agenticlybic/guiagent /app/.venv/bin/lybic-guiagent-grpc
+```
+> **Note**: The `-p 50051:50051` flag maps the container's gRPC port to your host machine.
+
+**2. Python Client Example**
+
+Once the service is running, you can interact with it using a gRPC client. Here is a Python example of how to send an instruction to the agent and stream its progress.
+
+First, ensure you have the necessary gRPC libraries and generated protobuf stubs:
+```sh
+# Install gRPC tools
+pip install grpcio grpcio-tools
+
+# Generate stubs from the .proto file
+python -m grpc_tools.protoc -Igui_agents/proto --python_out=gui_agents/proto/pb --grpc_python_out=gui_agents/proto/pb --pyi_out=gui_agents/proto/pb gui_agents/proto/agent.proto
+```
+
+Then, you can use the following script to communicate with the agent:
+
+```python
+import asyncio
+import grpc
+from gui_agents.proto.pb import agent_pb2, agent_pb2_grpc
+
+async def run_agent_instruction():
+    # Connect to the gRPC server
+    async with grpc.aio.insecure_channel('localhost:50051') as channel:
+        # Create a stub for the Agent service
+        stub = agent_pb2_grpc.AgentStub(channel)
+
+        # Create a request to run an instruction
+        request = agent_pb2.RunAgentInstructionRequest(
+            instruction="Open a calculator and compute 1 + 1"
+        )
+
+        print(f"Sending instruction: '{request.instruction}'")
+
+        # Call the RunAgentInstruction RPC and iterate over the stream of responses
+        try:
+            async for response in stub.RunAgentInstruction(request):
+                print(f"[{response.stage}] {response.message}")
+        except grpc.aio.AioRpcError as e:
+            print(f"An error occurred: {e.details()}")
+
+if __name__ == '__main__':
+    asyncio.run(run_agent_instruction())
+```
+
 ### VMware Configuration
 
 To use PyAutoGUI with VMware, you need to install [VMware Workstation Pro](https://www.vmware.com/products/desktop-hypervisor/workstation-and-fusion) (on Windows) and create a virtual machine. 

{lybic_guiagents-0.2.2 → lybic_guiagents-0.2.3}/gui_agents/__init__.py

@@ -37,7 +37,7 @@ from .agents.hardware_interface import HardwareInterface
 from .store.registry import Registry
 from .agents.global_state import GlobalState
 
-__version__ = "0.2.2"
+__version__ = "0.2.3"
 
 # Primary exports (what users should typically use)
 __all__ = [

{lybic_guiagents-0.2.2 → lybic_guiagents-0.2.3}/gui_agents/agents/Backend/LybicBackend.py

@@ -27,7 +27,7 @@ from gui_agents.agents.Backend.Backend import Backend
 
 # 导入官方Lybic SDK
 try:
-    from lybic import LybicClient, Sandbox, ComputerUse, dto
+    from lybic import LybicClient, Sandbox, ComputerUse, dto, LybicAuth
 except ImportError:
     raise ImportError(
         "Lybic Python SDK not found. Please install it with: pip install --upgrade lybic"
@@ -57,18 +57,22 @@ class LybicBackend(Backend):
                  precreate_sid: str = '',
                  **kwargs):
         """
-        初始化LybicBackend
-        
-        Args:
-            api_key: Lybic API密钥，如果为None则从环境变量LYBIC_API_KEY获取
-            org_id: Lybic组织ID，如果为None则从环境变量LYBIC_ORG_ID获取
-            endpoint: API端点，如果为None则从环境变量LYBIC_API_ENDPOINT获取
-            timeout: API请求超时时间
-            extra_headers: 额外的HTTP头
-            sandbox_opts: 创建沙盒时的额外选项
-            max_retries: 最大重试次数
-            precreate_sid: 预创建的沙盒ID，如果提供则不会创建新沙盒
-        """
+                 Initialize the LybicBackend, create and configure the Lybic SDK client, and ensure a sandbox is available.
+                 
+                 Parameters:
+                     api_key (Optional[str]): Lybic API key; if None the value is read from the LYBIC_API_KEY environment variable.
+                     org_id (Optional[str]): Lybic organization ID; if None the value is read from the LYBIC_ORG_ID environment variable.
+                     endpoint (Optional[str]): API endpoint; if None the value is read from LYBIC_API_ENDPOINT (default "https://api.lybic.cn").
+                     timeout (int): Request timeout in seconds for the SDK client.
+                     extra_headers (Optional[Dict[str, str]]): Additional HTTP headers to pass to the SDK via LybicAuth.
+                     sandbox_opts (Optional[Dict[str, Any]]): Options used when creating a new sandbox; LYBIC_MAX_LIFE_SECONDS is applied as the default for `maxLifeSeconds` if not provided.
+                     max_retries (int): Maximum number of retry attempts for action execution.
+                     precreate_sid (str): Pre-created sandbox ID to use; if empty, a new sandbox will be created.
+                 
+                 Raises:
+                     ValueError: If neither api_key nor org_id are provided (and not present in the corresponding environment variables).
+                     RuntimeError: If sandbox creation completes but no sandbox ID can be obtained from the SDK response.
+                 """
         self.loop = asyncio.new_event_loop()
         asyncio.set_event_loop(self.loop)
         
@@ -84,11 +88,13 @@ class LybicBackend(Backend):
         # 初始化SDK客户端（仅在有必要参数时）
         if self.api_key and self.org_id:
             self.client = LybicClient(
-                org_id=self.org_id,
-                api_key=self.api_key,
-                endpoint=self.endpoint,
+                LybicAuth(
+                    org_id=self.org_id,
+                    api_key=self.api_key,
+                    endpoint=self.endpoint,
+                    extra_headers=self.extra_headers or {}
+                ),
                 timeout=self.timeout,
-                extra_headers=self.extra_headers or {}
             )
         else:
             raise ValueError("LYBIC_API_KEY and LYBIC_ORG_ID are required. Please set them as environment variables or pass them as arguments.")
@@ -185,7 +191,7 @@ class LybicBackend(Backend):
             type="mouse:click",
             x=dto.PixelLength(type="px", value=act.x),
             y=dto.PixelLength(type="px", value=act.y),
-            button=1 if act.button == 0 else 2,  # 0=左键, 1=右键 -> 1=左键, 2=右键
+            button=act.button,
             holdKey=" ".join(act.holdKey) if act.holdKey else ""
         )
         
@@ -203,7 +209,7 @@ class LybicBackend(Backend):
             type="mouse:doubleClick",
             x=dto.PixelLength(type="px", value=act.x),
             y=dto.PixelLength(type="px", value=act.y),
-            button=1 if act.button == 0 else 2,
+            button=act.button,
             holdKey=" ".join(act.holdKey) if act.holdKey else ""
         )