collaboration_workflow.md

ReXia.AI CollaborationWorkflow Class

Overview

The CollaborationWorkflow class implements a workflow that utilizes team collaboration and tool usage for task completion. It's designed to be used within an Agent instance and extends the functionality of the BaseWorkflow class.

Class Attributes

• llm: The language model used by the workflow
• task: The task to be performed
• verbose: Flag for enabling verbose mode
• channel: The collaboration channel for communication
• team_work: The team worker component
• tool: The tool worker component

Methods

__init__(self, llm: Any, task: str, verbose: bool = False, max_attempts: int = 3) -> None

Initializes a CollaborationWorkflow instance with the following parameters:

• llm: The language model
• task: The assigned task
• verbose: Flag for verbose mode (default: False)
• max_attempts: Maximum attempts for model responses (default: 3)

_run_task(self) -> None

Internal method that executes the main workflow process:

1. Sets the task status to WORKING
2. Runs the tool component if tools are available
3. Runs the team work component
4. Sets the task status to COMPLETED

run(self) -> None

Public method to initiate the workflow execution.

Usage

To use the CollaborationWorkflow with an Agent:

from rexia_ai.workflows import CollaborationWorkflow
from rexia_ai.agent import Agent

llm = ...  # Your language model instance
task = "Collaborate on a research project and summarize findings."

agent = Agent(
    llm=llm,
    task=task,
    workflow=CollaborationWorkflow,
    verbose=True
)

result = agent.invoke()
print(result)

This workflow enhances task-solving capabilities by incorporating team collaboration and tool usage, making it suitable for complex tasks that require multiple steps or expertise.