ChatDev: Communicative Agents for Software Development
2023/07/16 by Chen Qian, Wei Liu, Qian, Chen +25 · 9 voices · 174 citations
Computer Science · #AI in Service Interactions #Software Engineering Research #Topic Modeling #cs.CL #cs.MA #cs.SE
paper · pdf · doi:10.48550/arxiv.2307.07924
openalex publication_date 2023/07/16 · openalex created_date 2023/07/20 · openalex updated_date 2026/07/28
Abstract
Software development is a complex task that necessitates cooperation among multiple members with diverse skills. Numerous studies used deep learning to improve specific phases in a waterfall model, such as design, coding, and testing. However, the deep learning model in each phase requires unique designs, leading to technical inconsistencies across various phases, which results in a fragmented and ineffective development process. In this paper, we introduce ChatDev, a chat-powered software development framework in which specialized agents driven by large language models (LLMs) are guided in what to communicate (via chat chain) and how to communicate (via communicative dehallucination). These agents actively contribute to the design, coding, and testing phases through unified language-based communication, with solutions derived from their multi-turn dialogues. We found their utilization of natural language is advantageous for system design, and communicating in programming language proves helpful in debugging. This paradigm demonstrates how linguistic communication facilitates multi-agent collaboration, establishing language as a unifying bridge for autonomous task-solving among LLM agents. The code and data are available at https://github.com/OpenBMB/ChatDev.
Cited by
- SPIRAL: Symbolic LLM Planning via Grounded and Reflective Search
- NERFIFY: A Multi-Agent Framework for Turning NeRF Papers into Code
- Monadic Context Engineering
- Agent2World: Learning to Generate Symbolic World Models via Adaptive Multi-Agent Feedback
- Focus Is All You Need: Adaptive Goal-aware Attention Orchestration for Multi-Agent Graph Systems
- Toward an Organizational Science of Multi-Agent LLM Systems: Decoupling Who, How, and Which Algorithm
- Agent Team Work Zone: An Automated, Persistent Workspace for Long-Lived Claude Code Agent Teams
- HeraSys: Collaborative Serving of Multiple LLM Workflows via Fine-Grained End-to-End Optimization
- TrajAudit: Automated Failure Diagnosis for Agentic Coding Systems
- Policy-Conditioned Policies for Multi-Agent Task Solving
- TrafficSimAgent: A Hierarchical Agent Framework for Autonomous Traffic Simulation with MCP Control
- LLM-Based Authoring of Agent-Based Narratives through Scene Descriptions
- A Multi-agent Text2SQL Framework using Small Language Models and Execution Feedback
- Re-opening open-source science through AI assisted development
- DUET: Agentic Design Understanding via Experimentation and Testing
- The Effect of Belief Boxes and Open-mindedness on Persuasion
- The Road of Adaptive AI for Precision in Cybersecurity
- ARCHER: Agentic Rule and Compliance Harness for Executable Regulations
- Mathematical Framing for Different Agent Strategies
- DataGovBench: Benchmarking LLM Agents for Real-World Data Governance Workflows
- SRPG: Semantically Reconstructed Privacy Guard for Zero-Trust Privacy in Educational Multi-Agent Systems
- Enhancing Automated Paper Reproduction via Prompt-Free Collaborative Agents
- RoleMotion: A Large-Scale Dataset towards Robust Scene-Specific Role-Playing Motion Synthesis with Fine-grained Descriptions
- A Flexible Multi-Agent LLM-Human Framework for Fast Human Validated Tool Building
- Hierarchical Decentralized Multi-Agent Coordination with Privacy-Preserving Knowledge Sharing: Extending AgentNet for Scalable Autonomous Systems
- LLM-Cave: A benchmark and light environment for large language models reasoning and decision-making system
- NOMAD: A Multi-Agent LLM System for UML Class Diagram Generation from Natural Language Requirements
- BAMAS: Structuring Budget-Aware Multi-Agent Systems
- The Consistency Critic: Correcting Inconsistencies in Generated Images via Reference-Guided Attentive Alignment
- Yo'City: Personalized and Boundless 3D Realistic City Scene Generation via Self-Critic Expansion
- Shadows in the Code: Exploring the Risks and Defenses of LLM-based Multi-Agent Software Development Systems
- End-to-End Automated Logging via Multi-Agent Framework
- LLM Assisted Coding with Metamorphic Specification Mutation Agent
- Paper2SysArch: Structure-Constrained System Architecture Generation from Scientific Papers
- SciEducator: Scientific Video Understanding and Educating via Deming-Cycle Multi-Agent System
- MURMUR: Using cross-user chatter to break collaborative language agents in groups
- Multi-Agent LLM Orchestration Achieves Deterministic, High-Quality Decision Support for Incident Response
- Designing LLM-based Multi-Agent Systems for Software Engineering Tasks: Quality Attributes, Design Patterns and Rationale
- Adaptive Multi-Agent Response Refinement in Conversational Systems
- From Natural Language to Certified H-infinity Controllers: Integrating LLM Agents with LMI-Based Synthesis
- On the Creativity of AI Agents
- Better Datasets Start From RefineLab: Automatic Optimization for High-Quality Dataset Refinement
- PRAGMA: A Profiling-Reasoned Multi-Agent Framework for Automatic Kernel Optimization
- Maestro: Learning to Collaborate via Conditional Listwise Policy Optimization for Multi-Agent LLMs
- Towards Realistic Project-Level Code Generation via Multi-Agent Collaboration and Semantic Architecture Modeling
- PublicAgent: Multi-Agent Design Principles From an LLM-Based Open Data Analysis Framework
- EvoDev: An Iterative Feature-Driven Framework for End-to-End Software Development with LLM-based Agents
- Optimal-Agent-Selection: State-Aware Routing Framework for Efficient Multi-Agent Collaboration
- MARS-SQL: A multi-agent reinforcement learning framework for Text-to-SQL
- How Focused Are LLMs? A Quantitative Study via Repetitive Deterministic Prediction Tasks
- Test-time Scaling of LLMs: A Survey from A Subproblem Structure Perspective
- Issue-Oriented Agent-Based Framework for Automated Review Comment Generation
- Validity Is What You Need
- A Research Roadmap for Augmenting Software Engineering Processes and Software Products with Generative AI
- CodeSpec: Dual Executable Specifications for Agentic Long-Horizon Feature Development
- Two Calls Beat Five Agents: Evaluating Multi-Agent Pipelines Against Self-Refinement for Local Language Models
- Enhancing Multi-Agent Communication through Attention Steering with Context Relevance
- CooperBench: Why Coding Agents Cannot be Your Teammates Yet
- Mitigating Hallucination in Large Language Models (LLMs): An Application-Oriented Survey on RAG, Reasoning, and Agentic Systems
- CodeCRDT: Observation-Driven Coordination for Multi-Agent LLM Code Generation
- TALM: Dynamic Tree-Structured Multi-Agent Framework with Long-Term Memory for Scalable Code Generation
- Collaborative LLM Agents for C4 Software Architecture Design Automation
- Integrating Machine Learning into Belief-Desire-Intention Agents: Current Advances and Open Challenges
- Knowledge-Guided Multi-Agent Framework for Application-Level Software Code Generation
- See, Think, Act: Online Shopper Behavior Simulation with VLM Agents
- When Your AI Agent Succumbs to Peer-Pressure: Studying Opinion-Change Dynamics of LLMs
- Enterprise Deep Research: Steerable Multi-Agent Deep Research for Enterprise Analytics
- Empowering Real-World: A Survey on the Technology, Practice, and Evaluation of LLM-driven Industry Agents
- Select-Then-Decompose: From Empirical Analysis to Adaptive Selection Strategy for Task Decomposition in Large Language Models
- TREAT: A Code LLMs Trustworthiness / Reliability Evaluation and Testing Framework
- ProtocolBench: Which LLM MultiAgent Protocol to Choose?
- MARSHAL: Incentivizing Multi-Agent Reasoning via Self-Play with Strategic LLMs
- Hi-Agent: Hierarchical Vision-Language Agents for Mobile Device Control
- Metacognitive Self-Correction for Multi-Agent System via Prototype-Guided Next-Execution Reconstruction
- Stop Reducing Responsibility in LLM-Powered Multi-Agent Systems to Local Alignment
- A2FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning
- Collaborative Shadows: Distributed Backdoor Attacks in LLM-Based Multi-Agent Systems
- Automating Structural Engineering Workflows with Large Language Model Agents
- KVComm: Enabling Efficient LLM Communication through Selective KV Sharing
- HyperAgent: Leveraging Hypergraphs for Topology Optimization in Multi-Agent Communication
- GraphTracer: Graph-Guided Failure Tracing in LLM Agents for Robust Multi-Turn Deep Search
- Merlin's Whisper: Enabling Efficient Reasoning in LLMs via Black-box Adversarial Prompting
- Testing and Enhancing Multi-Agent Systems for Robust Code Generation
- Sample-Efficient Online Learning in LM Agents via Hindsight Trajectory Rewriting
- SLEAN: Simple Lightweight Ensemble Analysis Network for Multi-Provider LLM Coordination: Design, Implementation, and Vibe Coding Bug Investigation Case Study
- Effective Strategies for Asynchronous Software Engineering Agents
- An Alternative Trajectory for Generative AI
- Humanoid Artificial Consciousness Designed with Large Language Model Based on Psychoanalysis and Personality Theory
- FOR-Prompting: From Objection to Revision via an Asymmetric Prompting Protocol
- Traceability and Accountability in Role-Specialized Multi-Agent LLM Pipelines
- AutoMLGen: Navigating Fine-Grained Optimization for Coding Agents
- AgentAsk: Multi-Agent Systems Need to Ask
- Codified Context: Infrastructure for AI Agents in a Complex Codebase
- From Simulation to Strategy: Automating Personalized Interaction Planning for Conversational Agents
- ARM: Discovering Agentic Reasoning Modules for Generalizable Multi-Agent Systems
- A Goal Without a Plan Is Just a Wish: Efficient and Effective Global Planner Training for Long-Horizon Agent Tasks
- From Agentification to Self-Evolving Agentic AI for Wireless Networks: Concepts, Approaches, and Future Research Directions
- MARS: Co-evolving Dual-System Deep Research via Multi-Agent Reinforcement Learning
- VortexPIA: Indirect Prompt Injection Attack against LLMs for Efficient Extraction of User Privacy
- Multi-Agent Code-Orchestrated Generation for Reliable Infrastructure-as-Code
- Adversarial Agent Collaboration for C to Rust Translation
- AutoMaAS: Self-Evolving Multi-Agent Architecture Search for Large Language Models
- Homophily-induced Emergence of Biased Structures in LLM-based Multi-Agent AI Systems
- Stochastic Self-Organization in Multi-Agent Systems
- Multi-LLM Orchestration for High-Quality Code Generation: Exploiting Complementary Model Strengths
- Flash-Searcher: Fast and Effective Web Agents via DAG-Based Parallel Execution
- A-MemGuard: A Proactive Defense Framework for LLM-Based Agent Memory
- CORRECT: COndensed eRror RECognition via knowledge Transfer in multi-agent systems
- Diagnose, Localize, Align: A Full-Stack Framework for Reliable LLM Multi-Agent Systems under Instruction Conflicts
- Generalized Multi-agent Social Simulation Framework
- RobustFlow: Towards Robust Agentic Workflow Generation
- Robust, Observable, and Evolvable Agentic Systems Engineering: A Principled Framework Validated via the Fairy GUI Agent
- A Fano-Style Accuracy Upper Bound for LLM Single-Pass Reasoning in Multi-Hop QA
- LLM Agent Meets Agentic AI: Can LLM Agents Simulate Customers to Evaluate Agentic-AI-based Shopping Assistants?
- What makes prompts a graph: necessary and sufficient conditions for prompt graph engineering
- FaithEyes: Towards Faithful Tool Use via Multi-Agent Process-Image Verification
- Scaling LLM-Driven Multi-Agent Systems: Design Principles and Architectural Scalability Analysis
- Agent Harness Distillation: Inference-Time Harness Extraction and Exploitation in Autonomous Multi-Agent Systems
- AgentRadio: Passive Awareness for Long-Horizon Multi-Agent Collaboration
- Auditing Emergent LLM-Agent Collaboration through Cooperation-Obligation Coupling
- Leveraging Trajectory Graphs for Pre-Execution Error Diagnosis in Agentic LLM Systems
- Escaping the Self-Confirmation Trap: An Execute-Distill-Verify Paradigm for Agentic Experience Learning
- AgentInit: Initializing LLM-based Multi-Agent Systems via Diversity and Expertise Orchestration for Effective and Efficient Collaboration
- Through the Lens of Human-Human Collaboration: A Configurable Research Platform for Exploring Human-Agent Collaboration
- Towards Transparent and Incentive-Compatible Collaboration in Decentralized LLM Multi-Agent Systems: A Blockchain-Driven Approach
- ChemOrch: Empowering LLMs with Chemical Intelligence via Synthetic Instructions
- RPG: A Repository Planning Graph for Unified and Scalable Codebase Generation
- The Psychology of Falsehood: A Human-Centric Survey of Misinformation Detection
- OnlineMate: An LLM-Based Multi-Agent Companion System for Cognitive Support in Online Learning
- (P)rior(D)yna(F)low: A Priori Dynamic Workflow Construction via Multi-Agent Collaboration
- MICA: Multi-Agent Industrial Coordination Assistant
- Who is Introducing the Failure? Automatically Attributing Failures of Multi-Agent Systems via Spectrum Analysis
- Evaluating Classical Software Process Models as Coordination Mechanisms for LLM-Based Software Generation
- Chinese Court Simulation with LLM-Based Agent System
- Crash Report Enhancement with Large Language Models: An Empirical Study
- From Language to Action: A Review of Large Language Models as Autonomous Agents and Tool Users
- From Legacy Fortran to Portable Kokkos: An Autonomous Agentic AI Workflow
- MOOM: Maintenance, Organization and Optimization of Memory in Ultra-Long Role-Playing Dialogues
- Auto-Slides: An Interactive Multi-Agent System for Creating and Customizing Research Presentations
- Abduct, Act, Predict: Scaffolding Causal Inference for Automated Failure Attribution in Multi-Agent Systems
- Automatic Failure Attribution and Critical Step Prediction Method for Multi-Agent Systems Based on Causal Inference
- Astra: A Multi-Agent System for GPU Kernel Performance Optimization
- Agentic Software Engineering: Foundational Pillars and a Research Roadmap
- Comp-X: On Defining an Interactive Learned Image Compression Paradigm With Expert-driven LLM Agent
- MAGneT: Coordinated Multi-Agent Generation of Synthetic Multi-Turn Mental Health Counseling Sessions
- RumorSphere: A Framework for Million-scale Agent-based Dynamic Simulation of Rumor Propagation
- Communicative Agents for Slideshow Storytelling Video Generation based on LLMs
- Towards Agentic OS: An LLM Agent Framework for Linux Schedulers
- PosterForest: Hierarchical Multi-Agent Collaboration for Scientific Poster Generation
- Principled Personas: Defining and Measuring the Intended Effects of Persona Prompting on Task Performance
- CompLex: Music Theory Lexicon Constructed by Autonomous Agents for Automatic Music Generation
- From Bits to Boardrooms: A Cutting-Edge Multi-Agent LLM Framework for Business Excellence
- STARec: An Efficient Agent Framework for Recommender Systems via Autonomous Deliberate Reasoning
- Requirements Development and Formalization for Reliable Code Generation: A Multi-Agent Vision
- Bias-Adjusted LLM Agents for Human-Like Decision-Making via Behavioral Economics
- Cognitive Agents Powered by Large Language Models for Agile Software Project Management
- Open-Universe Assistance Games
- Building and Measuring Trust between Large Language Models
- SafeSieve: From Heuristics to Experience in Progressive Pruning for LLM-based Multi-Agent Communication
- AI Agentic Programming: A Survey of Techniques, Challenges, and Opportunities
- Preacher: Paper-to-Video Agentic System
- OdysseyBench: Evaluating LLM Agents on Long-Horizon Complex Office Application Workflows
- Cowpox: Towards the Immunity of VLM-based Multi-Agent Systems
- DevNous: An LLM-Based Multi-Agent System for Grounding IT Project Management in Unstructured Conversation
- BrowseMaster: Towards Scalable Web Browsing via Tool-Augmented Programmatic Agent Pair
- BlindGuard: Safeguarding LLM-based Multi-Agent Systems under Unknown Attacks
- DRAMA: A Dynamic and Robust Allocation-based Multi-Agent System for Changing Environments
- StackPilot: Autonomous Function Agents for Scalable and Environment-Free Code Execution
- Tree-of-Reasoning: Towards Complex Medical Diagnosis via Multi-Agent Reasoning with Evidence Tree
- A Multi-Agent System for Complex Reasoning in Radiology Visual Question Answering
- A Survey on AgentOps: Categorization, Challenges, and Future Directions
- Large Language Model-based Data Science Agent: A Survey
- A Survey on Agent Workflow -- Status and Future
- WMAS: A Multi-Agent System Towards Intelligent and Customized Wireless Networks
Discussions
- link to the paper: arxiv.org/abs/2307.07924 [bsky, 4 points, 0 comments]
- Notes from the #QQI Masterclass with Danny Liu: ChatDev: Communicative Agents for Software Development arxiv.org/abs/2307.07924 [bsky, 2 points, 0 comments]
- Communicative Agents for Software Development [hn, 2 points, 0 comments]
- Communicative Agents for Software Development [hn, 2 points, 0 comments]
- Communicative LLM Agents for Software Development [hn, 1 points, 0 comments]
- Communicative Agents for Software Development [PDF] [hn, 1 points, 0 comments]
- I just stumbled upon this paper about communicative agents for software development based on LLM, which is used to simulate a waterfall development process. It sounds so crazy that I have to dig deepe [bsky, 1 points, 0 comments]
- arxiv.org/abs/2307.07924 [bsky, 0 points, 1 comments]
- Das Paper "Communicative Agents for Software Development" (arxiv.org/abs/2307.07924) und die Software (github.com/OpenBMB/ChatDev) sind frei verfügbar. So let's get started! 2/2 [bsky, 0 points, 0 comments]
Related