Skip to main content

openhands.memory.conversation_memory

ConversationMemory Objects​

class ConversationMemory()

Processes event history into a coherent conversation for the agent.

process_events​

def process_events(condensed_history: list[Event],
                   initial_messages: list[Message],
                   max_message_chars: int | None = None,
                   vision_is_active: bool = False,
                   enable_som_visual_browsing: bool = False) -> list[Message]

Process state history into a list of messages for the LLM.

Ensures that tool call actions are processed correctly in function calling mode.

Arguments:

  • state - The state containing the history of events to convert
  • condensed_history - The condensed list of events to process
  • initial_messages - The initial messages to include in the result
  • max_message_chars - The maximum number of characters in the content of an event included in the prompt to the LLM. Larger observations are truncated.
  • vision_is_active - Whether vision is active in the LLM. If True, image URLs will be included.
  • enable_som_visual_browsing - Whether to enable visual browsing for the SOM model.

process_initial_messages​

def process_initial_messages(with_caching: bool = False) -> list[Message]

Create the initial messages for the conversation.

apply_prompt_caching​

def apply_prompt_caching(messages: list[Message]) -> None

Applies caching breakpoints to the messages.

For new Anthropic API, we only need to mark the last user or tool message as cacheable.