RepoAgent: An LLM-Powered Open-Source Framework for Repository-level Code Documentation Generation

Published in

Conference on Empirical Methods in Natural Lang...(2024)

External Links:

Generate Graph

TL;DR

RepoAgent, a large language model powered open-source framework aimed at proactively generating, maintaining, and updating code documentation, is introduced, showing that RepoAgent excels in generating high-quality repository-level documentation.

Abstract

Generative models have demonstrated considerable potential in software engineering, particularly in tasks such as code generation and debugging. However, their utilization in the domain of code documentation generation remains underexplored. To this end, we introduce RepoAgent, a large language model powered open-source framework aimed at proactively generating, maintaining, and updating code documentation. Through both qualitative and quantitative evaluations, we have validated the effectiveness of our approach, showing that RepoAgent excels in generating high-quality repository-level documentation. The code and results are publicly accessible at https://github.com/OpenBMB/RepoAgent.

Authors

Maosong Sun

7 papers

Yankai Lin

9 papers

Yining Ye

3 papers

References43 items

Enhancing Open-Domain Task-Solving Capability of LLMs via Autonomous Tool Integration from GitHub

A Prompt Learning Framework for Source Code Summarization

ProAgent: From Robotic Process Automation to Agentic Process Automation

Lemur: Harmonizing Natural Language and Code for Language Agents

Code Llama: Open Foundation Models for Code

RepoAgent: An LLM-Powered Open-Source Framework for Repository-level Code Documentation Generation

Published in

Conference on Empirical Methods in Natural Lang...(2024)

External Links:

Generate Graph

TL;DR

Abstract

Authors

Maosong Sun

7 papers

Yankai Lin

9 papers

Yining Ye

3 papers

References43 items

Enhancing Open-Domain Task-Solving Capability of LLMs via Autonomous Tool Integration from GitHub

A Prompt Learning Framework for Source Code Summarization

ProAgent: From Robotic Process Automation to Agentic Process Automation

Lemur: Harmonizing Natural Language and Code for Language Agents

Code Llama: Open Foundation Models for Code

Shihao Liang

2 papers

Zhong Zhang

3 papers

Yingli Zhang

1 papers

Xiaoyin Che

1 papers

Zhiyuan Liu

1 papers

MetaGPT: Meta Programming for Multi-Agent Collaborative Framework

ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs

StarCoder: may the source be with you!

Tool Learning with Foundation Models

gDoc: Automatic Generation of Structured API Documentation

Automatic Code Documentation Generation Using GPT-3

CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesis

A Review on Source Code Documentation

A Survey of Automatic Source Code Summarization

Evaluating Large Language Models Trained on Code

Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer

A Novel Neural Source Code Representation Based on Abstract Syntax Tree

Measuring Program Comprehension: A Large-Scale Field Study with Professionals

Attention is All you Need

Summarizing Source Code using a Neural Attention Model

A Convolutional Attention Network for Extreme Summarization of Source Code

Improving automated source code summarization via an eye-tracking study of programmers

Automatic generation of natural language summaries for Java classes

On the Use of Automated Text Summarization Techniques for Summarizing Source Code

Towards automatically generating summary comments for Java methods

A study of the documentation essential to software maintenance

Programs, life cycles, and laws of software evolution

AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation Framework

Xagent: An autonomous agent for complex task solving

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

The Dependency Inversion Principle

Improving Language Understanding by Generative Pre-Training

Cost, benefits and quality of software development documentation: A systematic mapping

The study of the...

The ` RolePlaying ` class is used to simulate the chat session and is expected to have methods like ` init_chat ` and ` step `

2023. Communicative agents for software development

-Thefunction assumes that the ` chat_env ` has methods ` exist_employee ` to check for the existence of roles

2023. GPT-4 technical report

2024. De-bugBench: Evaluating debugging capability of large language models

2022. Chain-of-thought prompting elicits reasoning in large language models

if execute_command == ask_user.name: # HACK execute_result = ActionSuccessResult(outputs=user_input) agent.event_history.register_result(execute_result)

collaboration and exploring emergent behaviors

Field of Study

Computer Science

Journal Information

Name

ArXiv

Volume

abs/2005.00687

Venue Information

Name

Conference on Empirical Methods in Natural Language Processing

Type

conference

URL

https://www.aclweb.org/portal/emnlp

Alternate Names

Empir Method Nat Lang Process
Empirical Methods in Natural Language Processing
Conf Empir Method Nat Lang Process
EMNLP

TL;DR

Abstract

Authors

References43 items

Enhancing Open-Domain Task-Solving Capability of LLMs via Autonomous Tool Integration from GitHub

A Prompt Learning Framework for Source Code Summarization

ProAgent: From Robotic Process Automation to Agentic Process Automation

Lemur: Harmonizing Natural Language and Code for Language Agents

Code Llama: Open Foundation Models for Code

TL;DR

Abstract

Authors

References43 items

Enhancing Open-Domain Task-Solving Capability of LLMs via Autonomous Tool Integration from GitHub

A Prompt Learning Framework for Source Code Summarization

ProAgent: From Robotic Process Automation to Agentic Process Automation

Lemur: Harmonizing Natural Language and Code for Language Agents

Code Llama: Open Foundation Models for Code

MetaGPT: Meta Programming for Multi-Agent Collaborative Framework

ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs

StarCoder: may the source be with you!

Tool Learning with Foundation Models

gDoc: Automatic Generation of Structured API Documentation

Automatic Code Documentation Generation Using GPT-3

CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesis

A Review on Source Code Documentation

A Survey of Automatic Source Code Summarization

Evaluating Large Language Models Trained on Code

Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer

A Novel Neural Source Code Representation Based on Abstract Syntax Tree

Measuring Program Comprehension: A Large-Scale Field Study with Professionals

Attention is All you Need

Summarizing Source Code using a Neural Attention Model

A Convolutional Attention Network for Extreme Summarization of Source Code

Improving automated source code summarization via an eye-tracking study of programmers

Automatic generation of natural language summaries for Java classes

On the Use of Automated Text Summarization Techniques for Summarizing Source Code

Towards automatically generating summary comments for Java methods

A study of the documentation essential to software maintenance

Programs, life cycles, and laws of software evolution

AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation Framework

Xagent: An autonomous agent for complex task solving

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

The Dependency Inversion Principle

Improving Language Understanding by Generative Pre-Training

Cost, benefits and quality of software development documentation: A systematic mapping

The study of the...

The ` RolePlaying ` class is used to simulate the chat session and is expected to have methods like ` init_chat ` and ` step `

2023. Communicative agents for software development

-Thefunction assumes that the ` chat_env ` has methods ` exist_employee ` to check for the existence of roles

2023. GPT-4 technical report

2024. De-bugBench: Evaluating debugging capability of large language models

2022. Chain-of-thought prompting elicits reasoning in large language models

if execute_command == ask_user.__name__: # HACK execute_result = ActionSuccessResult(outputs=user_input) agent.event_history.register_result(execute_result)

2024. AgentVerse

collaboration and exploring emergent behaviors

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names

if execute_command == ask_user.name: # HACK execute_result = ActionSuccessResult(outputs=user_input) agent.event_history.register_result(execute_result)