Skip to main content
The ConversimpleAgent provides several callback methods that you can override in your agent class to react to different events during a conversation.

​
Callback Methods

​
on_conversation_started(conversation_id: str)

Called when a new conversation starts. 
def on_conversation_started(self, conversation_id: str) -> None:
    print(f"Conversation started: {conversation_id}")
​
conversation_id
str
required
The unique identifier for the started conversation.

​
on_conversation_ended(conversation_id: str)

Called when a conversation ends. 
def on_conversation_ended(self, conversation_id: str) -> None:
    print(f"Conversation ended: {conversation_id}")
​
conversation_id
str
required
The unique identifier for the ended conversation.

​
on_tool_called(tool_call)

Called when the platform requests your agent to execute a tool. 
def on_tool_called(self, tool_call) -> None:
    print(f"Executing tool: {tool_call.tool_name}")
​
tool_call
conversimple.tools.ToolCall
required
An object containing details about the tool call, including call_id, tool_name, arguments, and conversation_id.

​
on_tool_completed(call_id: str, result: Any)

Called after a tool execution has successfully completed and its result has been sent to the platform. 
def on_tool_completed(self, call_id: str, result: Any) -> None:
    print(f"Tool {call_id} completed with result: {result}")
​
call_id
str
required
The unique identifier of the tool call.
​
result
Any
required
The result returned by the executed tool.

​
on_error(error_type: str, message: str, details: dict)

Called when an error occurs within the agent or is reported by the platform. 
def on_error(self, error_type: str, message: str, details: dict) -> None:
    print(f"Error ({error_type}): {message}")
​
error_type
str
required
The type of error (e.g., connection_error, tool_execution_error).
​
message
str
required
A descriptive error message.
​
details
dict
required
A dictionary containing additional error details.

​
on_config_update(config: Dict[str, Any])

Called when the agent receives a configuration update from the platform. 
def on_config_update(self, config: Dict[str, Any]) -> None:
    print(f"Configuration updated: {config}")
​
config
Dict[str, Any]
required
A dictionary containing the updated configuration.