Paper - ChatDev: Communicative Agents for Software Development

Overview

ChatDev is a research framework for software development that uses multiple large language model (LLM) driven agents communicating via chat. Each agent specializes in a development phase such as design, coding, or testing, aiming to maintain consistency across phases. The approach seeks to replace fragmented deep learning models with a unified, chat-based collaboration paradigm.

Best for

Best for
Researchers exploring multi-agent LLM frameworks for autonomous software development

Use cases

• Simulating multi-agent collaborative code generation
• Prototyping an LLM-driven software development lifecycle
• Experimenting with autonomous agent coordination for programming tasks

Notes

ChatDev is a research framework for software development that uses multiple large language model (LLM) driven agents communicating via chat. Each agent specializes in a development phase such as design, coding, or testing, aiming to maintain consistency across phases. The approach seeks to replace fragmented deep learning models with a unified, chat-based collaboration paradigm.

Use cases

• Simulating multi-agent collaborative code generation
• Prototyping an LLM-driven software development lifecycle
• Experimenting with autonomous agent coordination for programming tasks

Pros

• Reduces fragmentation by using a single chat-based framework across phases
• Leverages natural language interaction for intuitive agent collaboration
• Novel approach that highlights limitations of current deep learning models in development

Cons

• Currently a research paper, not a production-ready tool
• Performance depends heavily on underlying LLM quality and prompt engineering
• Scalability and reliability of multi-agent systems in complex projects remain unproven

Indexed from awesome-ai-agents and enriched against its public facts.