Merge branch 'main' into kevin

Author: SmartManoj

config.template.toml

@@ -319,6 +319,10 @@ llm_config = 'gpt3'
 # Additional Docker runtime kwargs
 #docker_runtime_kwargs = {}
 
+# Specific port to use for VSCode. If not set, a random port will be chosen.
+# Useful when deploying OpenHands in a remote machine where you need to expose a specific port.
+#vscode_port = 41234
+
 #################################### Security ###################################
 # Configuration for security features
 ##############################################################################

docs/modules/usage/troubleshooting/troubleshooting.md

@@ -4,6 +4,38 @@
 OpenHands only supports Windows via WSL. Please be sure to run all commands inside your WSL terminal.
 :::
 
+### Unable to access VS Code tab via local IP
+
+**Description**
+
+When accessing OpenHands through a non-localhost URL (such as a LAN IP address), the VS Code tab shows a "Forbidden" error, while other parts of the UI work fine.
+
+**Resolution**
+
+This happens because VS Code runs on a random high port that may not be exposed or accessible from other machines. To fix this:
+
+1. Set a specific port for VS Code using the `SANDBOX_VSCODE_PORT` environment variable:
+   ```bash
+   docker run -it --rm \
+       -e SANDBOX_VSCODE_PORT=41234 \
+       -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:latest \
+       -v /var/run/docker.sock:/var/run/docker.sock \
+       -v ~/.openhands-state:/.openhands-state \
+       -p 3000:3000 \
+       -p 41234:41234 \
+       --add-host host.docker.internal:host-gateway \
+       --name openhands-app \
+       docker.all-hands.dev/all-hands-ai/openhands:latest
+   ```
+
+2. Make sure to expose the same port with `-p 41234:41234` in your Docker command.
+
+3. Alternatively, you can set this in your `config.toml` file:
+   ```toml
+   [sandbox]
+   vscode_port = 41234
+   ```
+
 ### Launch docker client failed
 
 **Description**

openhands/agenthub/codeact_agent/codeact_agent.py

@@ -1,8 +1,12 @@
 import copy
 import os
 from collections import deque
+from typing import TYPE_CHECKING
 
-from litellm import ChatCompletionToolParam
+if TYPE_CHECKING:
+    from litellm import ChatCompletionToolParam
+    from openhands.events.action import Action
+    from openhands.llm.llm import ModelResponse
 
 import openhands.agenthub.codeact_agent.function_calling as codeact_function_calling
 from openhands.agenthub.codeact_agent.tools.bash import create_cmd_run_tool
@@ -20,7 +24,7 @@
 from openhands.core.config import AgentConfig
 from openhands.core.logger import openhands_logger as logger
 from openhands.core.message import Message
-from openhands.events.action import Action, AgentFinishAction, MessageAction
+from openhands.events.action import AgentFinishAction, MessageAction
 from openhands.events.event import Event
 from openhands.llm.llm import LLM
 from openhands.memory.condenser import Condenser
@@ -75,23 +79,26 @@ def __init__(
         - config (AgentConfig): The configuration for this agent
         """
         super().__init__(llm, config)
-        self.pending_actions: deque[Action] = deque()
+        self.pending_actions: deque['Action'] = deque()
         self.reset()
         self.tools = self._get_tools()
 
-        self.prompt_manager = PromptManager(
-            prompt_dir=os.path.join(os.path.dirname(__file__), 'prompts'),
-        )
-
         # Create a ConversationMemory instance
         self.conversation_memory = ConversationMemory(self.config, self.prompt_manager)
 
         self.condenser = Condenser.from_config(self.config.condenser)
         logger.debug(f'Using condenser: {type(self.condenser)}')
 
-        self.response_to_actions_fn = codeact_function_calling.response_to_actions
+    @property
+    def prompt_manager(self) -> PromptManager:
+        if self._prompt_manager is None:
+            self._prompt_manager = PromptManager(
+                prompt_dir=os.path.join(os.path.dirname(__file__), 'prompts'),
+            )
 
-    def _get_tools(self) -> list[ChatCompletionToolParam]:
+        return self._prompt_manager
+
+    def _get_tools(self) -> list['ChatCompletionToolParam']:
         # For these models, we use short tool descriptions ( < 1024 tokens)
         # to avoid hitting the OpenAI token limit for tool descriptions.
         SHORT_TOOL_DESCRIPTION_LLM_SUBSTRS = ['gpt-', 'o3', 'o1', 'o4']
@@ -130,7 +137,7 @@ def reset(self) -> None:
         super().reset()
         self.pending_actions.clear()
 
-    def step(self, state: State) -> Action:
+    def step(self, state: State) -> 'Action':
         """Performs one step using the CodeAct Agent.
 
         This includes gathering info on previous steps and prompting the model to make a command to execute.
@@ -198,9 +205,7 @@ def step(self, state: State) -> Action:
         params['extra_body'] = {'metadata': state.to_llm_metadata(agent_name=self.name)}
         response = self.llm.completion(**params)
         logger.debug(f'Response from LLM: {response}')
-        actions = self.response_to_actions_fn(
-            response, mcp_tool_names=list(self.mcp_tools.keys())
-        )
+        actions = self.response_to_actions(response)
         logger.debug(f'Actions after response_to_actions: {actions}')
         for action in actions:
             self.pending_actions.append(action)
@@ -273,4 +278,9 @@ def _get_messages(
         if self.llm.is_caching_prompt_active():
             self.conversation_memory.apply_prompt_caching(messages)
 
-        return messages
+        return messages
+
+    def response_to_actions(self, response: 'ModelResponse') -> list['Action']:
+        return codeact_function_calling.response_to_actions(
+            response, mcp_tool_names=list(self.mcp_tools.keys())
+        )

openhands/agenthub/readonly_agent/readonly_agent.py

@@ -4,6 +4,13 @@
 
 import os
 
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from litellm import ChatCompletionToolParam
+    from openhands.events.action import Action
+    from openhands.llm.llm import ModelResponse
+
 from openhands.agenthub.codeact_agent.codeact_agent import CodeActAgent
 from openhands.agenthub.readonly_agent import (
     function_calling as readonly_function_calling,
@@ -41,24 +48,27 @@ def __init__(
         - llm (LLM): The llm to be used by this agent
         - config (AgentConfig): The configuration for this agent
         """
-        # Initialize the CodeActAgent class but we'll override some of its behavior
+        # Initialize the CodeActAgent class; some of it is overridden with class methods
         super().__init__(llm, config)
 
-        # Override the tools to only include read-only tools
-        # Get the read-only tools from our own function_calling module
-        self.tools = readonly_function_calling.get_tools()
-
-        # Set up our own prompt manager
-        self.prompt_manager = PromptManager(
-            prompt_dir=os.path.join(os.path.dirname(__file__), 'prompts'),
-        )
-
-        self.response_to_actions_fn = readonly_function_calling.response_to_actions
-
         logger.debug(
             f"TOOLS loaded for ReadOnlyAgent: {', '.join([tool.get('function').get('name') for tool in self.tools])}"
         )
 
+    @property
+    def prompt_manager(self) -> PromptManager:
+        # Set up our own prompt manager
+        if self._prompt_manager is None:
+            self._prompt_manager = PromptManager(
+                prompt_dir=os.path.join(os.path.dirname(__file__), 'prompts'),
+            )
+        return self._prompt_manager
+
+    def _get_tools(self) -> list['ChatCompletionToolParam']:
+        # Override the tools to only include read-only tools
+        # Get the read-only tools from our own function_calling module
+        return readonly_function_calling.get_tools()
+
     def set_mcp_tools(self, mcp_tools: list[dict]) -> None:
         """Sets the list of MCP tools for the agent.
 
@@ -68,3 +78,8 @@ def set_mcp_tools(self, mcp_tools: list[dict]) -> None:
         logger.warning(
             'ReadOnlyAgent does not support MCP tools. MCP tools will be ignored by the agent.'
         )
+
+    def response_to_actions(self, response: 'ModelResponse') -> list['Action']:
+        return readonly_function_calling.response_to_actions(
+            response, mcp_tool_names=list(self.mcp_tools.keys())
+        )

openhands/core/cli_commands.py openhands/cli/commands.pyopenhands/core/cli_commands.py renamed to openhands/cli/commands.py

@@ -5,20 +5,20 @@
 from prompt_toolkit.shortcuts import clear, print_container
 from prompt_toolkit.widgets import Frame, TextArea
 
-from openhands.core.cli_settings import (
+from openhands.cli.settings import (
     display_settings,
     modify_llm_settings_advanced,
     modify_llm_settings_basic,
 )
-from openhands.core.cli_tui import (
+from openhands.cli.tui import (
     COLOR_GREY,
     UsageMetrics,
     cli_confirm,
     display_help,
     display_shutdown_message,
     display_status,
 )
-from openhands.core.cli_utils import (
+from openhands.cli.utils import (
     add_local_config_trusted_dir,
     get_local_config_trusted_dirs,
     read_file,

openhands/core/cli.py openhands/cli/main.pyopenhands/core/cli.py renamed to openhands/cli/main.py

@@ -7,13 +7,11 @@
 from prompt_toolkit.shortcuts import clear
 
 import openhands.agenthub  # noqa F401 (we import this to get the agents registered)
-from openhands.controller import AgentController
-from openhands.controller.agent import Agent
-from openhands.core.cli_commands import (
+from openhands.cli.commands import (
     check_folder_security_agreement,
     handle_commands,
 )
-from openhands.core.cli_tui import (
+from openhands.cli.tui import (
     UsageMetrics,
     display_agent_running_message,
     display_banner,
@@ -26,9 +24,11 @@
     read_confirmation_input,
     read_prompt_input,
 )
-from openhands.core.cli_utils import (
+from openhands.cli.utils import (
     update_usage_metrics,
 )
+from openhands.controller import AgentController
+from openhands.controller.agent import Agent
 from openhands.core.config import (
     AppConfig,
     parse_arguments,

openhands/core/cli_settings.py openhands/cli/settings.pyopenhands/core/cli_settings.py renamed to openhands/cli/settings.py

@@ -5,19 +5,19 @@
 from prompt_toolkit.widgets import Frame, TextArea
 from pydantic import SecretStr
 
-from openhands.controller.agent import Agent
-from openhands.core.cli_tui import (
+from openhands.cli.tui import (
     COLOR_GREY,
     UserCancelledError,
     cli_confirm,
     kb_cancel,
 )
-from openhands.core.cli_utils import (
+from openhands.cli.utils import (
     VERIFIED_ANTHROPIC_MODELS,
     VERIFIED_OPENAI_MODELS,
     VERIFIED_PROVIDERS,
     organize_models_and_providers,
 )
+from openhands.controller.agent import Agent
 from openhands.core.config import AppConfig
 from openhands.core.config.condenser_config import NoOpCondenserConfig
 from openhands.core.config.utils import OH_DEFAULT_AGENT

openhands/core/cli_tui.py openhands/cli/tui.pyopenhands/core/cli_tui.py renamed to openhands/cli/tui.py

@@ -3,7 +3,7 @@
 
 import toml
 
-from openhands.core.cli_tui import (
+from openhands.cli.tui import (
     UsageMetrics,
 )
 from openhands.events.event import Event

openhands/core/cli_utils.py openhands/cli/utils.pyopenhands/core/cli_utils.py renamed to openhands/cli/utils.py

@@ -8,6 +8,7 @@
     from openhands.core.config import AgentConfig
     from openhands.events.action import Action
     from openhands.events.action.message import SystemMessageAction
+    from openhands.utils.prompt import PromptManager
 from litellm import ChatCompletionToolParam
 
 from openhands.core.exceptions import (
@@ -22,9 +23,6 @@
 from openhands.llm.llm import LLM
 from openhands.runtime.plugins import PluginRequirement
 
-if TYPE_CHECKING:
-    from openhands.utils.prompt import PromptManager
-
 
 class Agent(ABC):
     DEPRECATED = False
@@ -46,11 +44,17 @@ def __init__(
         self.llm = llm
         self.config = config
         self._complete = False
-        self.prompt_manager: 'PromptManager' | None = None
+        self._prompt_manager: 'PromptManager' | None = None
         self.event_stream: 'EventStream' | None = None
         self.mcp_tools: dict[str, ChatCompletionToolParam] = {}
         self.tools: list = []
 
+    @property
+    def prompt_manager(self) -> 'PromptManager':
+        if self._prompt_manager is None:
+            raise ValueError(f'Prompt manager not initialized for agent {self.name}')
+        return self._prompt_manager
+
     def get_system_message(self) -> 'SystemMessageAction | None':
         """
         Returns a SystemMessageAction containing the system message and tools.