README.md

Table of Contents

• Structure of PlanWriterFlow
• run_planwriter
• PlanWriterAskUserFlow
  • PlanWriterAskUserFlow
    • run
• PlanWriterFlow
  • PlanWriterFlow
    • detect_finish_or_continue
    • run
• __init__
• PlanWriterCtrlFlow
  • PlanWriterCtrlFlow
    • __init__
    • instantiate_from_config
    • run

Structure of PlanWriterFlow

               goal
                |
                v
        +---------------+
        |  Controller   | --------<<<<-----------+
        +---------------+                        |
                |                                |
                | (command, command args)        |
                |                                |
                v                                |
        +------------------+                     |
        |   Executor       |  Each branch is an  |
        | (Tree Structure) |  executor           |
        +------------------+                     |
                |                                ^
                | (summary)                      |
                |                                |
                v                                |             
                |                                |
                +-> goes back to the Controller>-+

Structure of the Executors:

         +-------------------+
         |   Branching       |
         |    Executor       |
         +-------------------+
        /                    \
       /                      \
      /                        \
     /                          \
write_plan                    ask_user

About the branches:

• ask_user: Ask user for info / confirmation, etc.
• write_plan: Generates plan (user edit is allowed) and fetches user feedback.

How it works: Controller calls write_plan until user is satisfied in the feedback, finish.

run_planwriter

PlanWriterAskUserFlow

PlanWriterAskUserFlow Objects

class PlanWriterAskUserFlow(HumanStandardInputFlow)

Refer to: https://huggingface.co/aiflows/ExtendLibraryFlowModule/blob/main/ExtLibAskUserFlow.py This flow is used to ask the user a question and get a response.

Input Interface:

• question

Output Interface:

• feedback
• plan

Configuration Parameters:

• query_message_prompt_template: The message template to prompt the user for input.
• request_multi_line_input_flag: Whether to request multi-line input from the user.
• end_of_input_string: The string to enter to indicate the end of input.

run

def run(input_data: Dict[str, Any]) -> Dict[str, Any]

Run the flow module.

Arguments:

• input_data (Dict[str, Any]): The input data to the flow module.

Returns:

Dict[str, Any]: The output data from the flow module.

PlanWriterFlow

PlanWriterFlow Objects

class PlanWriterFlow(ContentWriterFlow)

This flow inherits from ContentWriterFlow. In the subflow of the executor, we specify the InteractivePlanGenFlow (https://huggingface.co/aiflows/InteractivePlanGenFlowModule)

Input Interface:

• goal

Output Interface:

• plan
• result
• summary
• status

Configuration Parameters:

• Also refer to superclass ContentWriterFlow (https://huggingface.co/aiflows/ContentWriterFlowModule)
• input_interface: the input interface of the flow
• output_interface: the output interface of the flow
• subflows_config: the configuration of the subflows
• early_exit_key: the key in the flow state that indicates the early exit condition
• topology: the topology of the flow

detect_finish_or_continue

@CircularFlow.output_msg_payload_processor
def detect_finish_or_continue(output_payload: Dict[str, Any],
                              src_flow) -> Dict[str, Any]

This function is called when the subflow finishes running.

configured in the topology of the subflow config.

Arguments:

• output_payload (Dict[str, Any]): the output payload of the subflow
• src_flow (Flow): the subflow that generates the output payload

Returns:

Dict[str, Any]: the processed output payload

run

def run(input_data: Dict[str, Any]) -> Dict[str, Any]

This function runs the flow.

Arguments:

• input_data (Dict[str, Any]): the input data of the flow

Returns:

Dict[str, Any]: the output data of the flow

__init__

PlanWriterCtrlFlow

PlanWriterCtrlFlow Objects

class PlanWriterCtrlFlow(ChatAtomicFlow)

Refer to: https://huggingface.co/aiflows/JarvisFlowModule/blob/main/Controller_JarvisFlow.py This flow is a controller flow that controls the PlanWriterFlow.

Input Interface Non Initialized:

• goal: str

Input Interface Initialized:

• feedback: str
• goal: str
• plan: str

Output Interface:

• command: str
• command_args: Dict[str, Any]

Configuration Parameters:

• input_interface_non_initialized: List[str] = ["goal"]
• input_interface_initialized: List[str] = ["feedback", "goal", "plan"]
• output_interface: List[str] = ["command", "command_args"]
• backend: Dict[str, Any] : backend of the LLM
• commands: List[Dict[str, Any]] : commands that the LLM can execute
• system_message_prompt_template: str : the template of the system message prompt
• init_human_message_prompt_template: str : the template of the initial human message prompt
• human_message_prompt_template: str : the template of the human message prompt
• previous_messages: Dict[str, Any] : the previous messages of the conversation (sliding window)

__init__

def __init__(commands: List[Command], **kwargs)

This function initializes the flow.

Arguments:

• commands (List[Command]): List[Command] : commands that the LLM can execute
• kwargs (Dict[str, Any]): other parameters

instantiate_from_config

@classmethod
def instantiate_from_config(cls, config)

This function instantiates the flow from the config.

Arguments:

• config (Dict[str, Any]): the config of the flow

Returns:

ChatAtomicFlow: the instantiated flow

run

def run(input_data: Dict[str, Any]) -> Dict[str, Any]

This function runs the flow.

Arguments:

• input_data (Dict[str, Any]): the input data

Returns:

Dict[str, Any]: the output of the flow