Skip to content

mrnust/MAS_with_LLMs

Repository files navigation

Project 1: LLM-Powered Multi-Agent Code Management System

Overview

This project is a Code Generation and Management Tool that leverages a multi-agent system powered by the Llama 3 language model (LLM). The tool is designed to automate and streamline the process of code generation, optimization, error resolution, and reporting. It simulates a team of agents working together to efficiently produce high-quality code based on user-provided requirements.

Features

  • CEO Agent: Oversees the entire project, sets objectives, and reviews the final report.
  • Manager Agent: Manages the workflow of the team, assigns tasks, and ensures task completion.
  • Code Generation Agent: Generates the initial code based on the given requirements.
  • Code Optimization Agent: Optimizes the generated code by applying best practices.
  • Error Resolution Agent: Identifies and resolves any errors in the code.
  • Reporting Agent: Generates a comprehensive report on the project's progress, including code quality and errors.

Project 2: LLM-Powered Multi-Agent Content Generation

Overview

This project is a Content Generation Tool that utilizes a multi-agent system powered by the Llama 3 language model (LLM). The tool automates the process of researching, generating, and simplifying content based on a user-provided topic. It features agents specialized in different aspects of content creation, working together to produce high-quality, accessible content.

Features

  • Research Agent: Gathers information from Wikipedia based on the provided topic and generates a draft.
  • Content Generator Agent: Produces detailed content or an article based on the draft.
  • Simplifier Agent: Simplifies the generated content to make it more accessible and easier to understand.