# RPG: A Repository Planning Graph for Unified and Scalable Codebase Generation

arXiv: https://arxiv.org/abs/2509.16198 · HTML: https://mlsanity.com/paper/2509.16198/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- Builds on 32 of its 47 references that are in the map.
- First posted on arXiv 19 Sep 2025.

## Builds on (most cited first, 30 of 32)

- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation](https://mlsanity.com/paper/2308.08155/) (2023)
- [MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework](https://mlsanity.com/paper/2308.00352/) (2023)
- [Qwen2.5-Coder Technical Report](https://mlsanity.com/paper/2409.12186/) (2024)
- [OpenHands: An Open Platform for AI Software Developers as Generalist Agents](https://mlsanity.com/paper/2407.16741/) (2024)
- [ChatDev: Communicative Agents for Software Development](https://mlsanity.com/paper/2307.07924/) (2023)
- [Graph of Thoughts: Solving Elaborate Problems with Large Language Models](https://mlsanity.com/paper/2308.09687/) (2023)
- [DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligence](https://mlsanity.com/paper/2406.11931/) (2024)
- [PaperBench: Evaluating AI's Ability to Replicate AI Research](https://mlsanity.com/paper/2504.01848/) (2025)
- [AdaPlanner: Adaptive Planning from Feedback with Language Models](https://mlsanity.com/paper/2305.16653/) (2023)
- [ACECODER: Acing Coder RL via Automated Test-Case Synthesis](https://mlsanity.com/paper/2502.01718/) (2025)
- [On the Planning Abilities of Large Language Models (A Critical Investigation with a Proposed Benchmark)](https://mlsanity.com/paper/2302.06706/) (2023)
- [Planning In Natural Language Improves LLM Search For Code Generation](https://mlsanity.com/paper/2409.03733/) (2024)
- [Paper2Code: Automating Code Generation from Scientific Papers in Machine Learning](https://mlsanity.com/paper/2504.17192/) (2025)
- [DA-Code: Agent Data Science Code Generation Benchmark for Large Language Models](https://mlsanity.com/paper/2410.07331/) (2024)
- [FEA-Bench: A Benchmark for Evaluating Repository-Level Code Generation for Feature Implementation](https://mlsanity.com/paper/2503.06680/) (2025)
- [Commit0: Library Generation from Scratch](https://mlsanity.com/paper/2412.01769/) (2024)
- [Effective Test Generation Using Pre-trained Large Language Models and Mutation Testing](https://mlsanity.com/paper/2308.16557/) (2023)
- [rStar-Coder: Scaling Competitive Code Reasoning with a Large-Scale Verified Dataset](https://mlsanity.com/paper/2505.21297/) (2025)
- [RefactorBench: Evaluating Stateful Reasoning in Language Agents Through Code](https://mlsanity.com/paper/2503.07832/) (2025)
- [Code Graph Model (CGM): A Graph-Integrated Large Language Model for Repository-Level Software Engineering Tasks](https://mlsanity.com/paper/2505.16901/) (2025)
- [EpiCoder: Encompassing Diversity and Complexity in Code Generation](https://mlsanity.com/paper/2501.04694/) (2025)
- [AutoP2C: An LLM-Based Agent Framework for Code Repository Generation from Multimodal Content in Academic Papers](https://mlsanity.com/paper/2504.20115/) (2025)
- [Enhancing LLM Code Generation: A Systematic Evaluation of Multi-Agent Collaboration and Runtime Debugging for Improved Accuracy, Reliability, and Latency](https://mlsanity.com/paper/2505.02133/) (2025)
- [Guided Code Generation with LLMs: A Multi-Agent Framework for Complex Code Tasks](https://mlsanity.com/paper/2501.06625/) (2025)
- [AdaCoder: An Adaptive Planning and Multi-Agent Framework for Function-Level Code Generation](https://mlsanity.com/paper/2504.04220/) (2025)
- [How Far Are LLMs from Symbolic Planners? An NLP-Based Perspective](https://mlsanity.com/paper/2508.01300/) (2025)

## Built on by (most cited first, 19 of 19)

- [Code as Agent Harness](https://mlsanity.com/paper/2605.18747/) (2026)
- [Retrieval-Augmented Code Generation: A Survey with Focus on Repository-Level Approaches](https://mlsanity.com/paper/2510.04905/) (2025)
- [TestExplora: Benchmarking LLMs for Proactive Bug Discovery via Repository-Level Test Generation](https://mlsanity.com/paper/2602.10471/) (2026)
- [NERFIFY: A Multi-Agent Framework for Turning NeRF Papers into Code](https://mlsanity.com/paper/2603.00805/) (2026)
- [Constraint Decay: The Fragility of LLM Agents in Backend Code Generation](https://mlsanity.com/paper/2605.06445/) (2026)
- [Scientific reasoning does not reliably translate into scientific forecasting in frontier AI](https://mlsanity.com/paper/2605.22681/) (2026)
- [Repo0: Design-Driven Zero-to-All Code Generation](https://mlsanity.com/paper/2608.19854/) (2026)
- [Towards Iterative End-to-End Software Development: A Feature-Driven Multi-Agent Framework](https://mlsanity.com/paper/2511.02399/) (2025)
- [MindForge: Teaching Small Language Models Whole-Life-Cycle Software Engineering via Source-Free Program Synthesis](https://mlsanity.com/paper/2607.27146/) (2026)
- [Enhancing Automated Paper Reproduction via Prompt-Free Collaborative Agents](https://mlsanity.com/paper/2512.02812/) (2025)
- [X-Coder: Advancing Competitive Programming with Synthetic Tasks, Solutions, and Tests](https://mlsanity.com/paper/2601.06953/) (2026)
- [ReCUBE: Evaluating Repository-Level Context Utilization in Code Generation](https://mlsanity.com/paper/2603.25770/) (2026)
- [Persistent Cross-Attempt State Optimization for Repository-Level Code Generation](https://mlsanity.com/paper/2604.03632/) (2026)
- [TraceDev: A Traceability-Driven Multi-agent Framework for Requirement-to-Code Development](https://mlsanity.com/paper/2607.18886/) (2026)
- [Documentation-Guided Agentic Codebase Migration from C to Rust](https://mlsanity.com/paper/2605.14634/) (2026)
- [DPIAgent: Divide, Protocol, Isolate for Agentic Reproduction Test Generation](https://mlsanity.com/paper/2608.23341/) (2026)
- [The Thousand-Graph Hypothesis: A Testable Hypothesis of Task-Conditioned Relation Materialization in Repository-Level Code Reasoning](https://mlsanity.com/paper/2608.26602/) (2026)
- [SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving](https://mlsanity.com/paper/2609.26777/) (2026)
- [Don't Guess, Just Ask: Evaluating Interactive Requirement Clarification for Repository-Level Software Development](https://mlsanity.com/paper/2610.13022/) (2026)

## Abstract (arXiv)

Large language models excel at generating individual functions or single files of code, yet generating complete repositories from scratch remains a fundamental challenge. This capability is key to building coherent software systems from high-level specifications and realizing the full potential of automated code generation. The process requires planning at two levels: deciding what features and modules to build (proposal stage) and defining their implementation details (implementation stage). Current approaches rely on natural language planning, which often produces unclear specifications, misaligned components, and brittle designs due to its inherent ambiguity and lack of structure. To address these limitations, we introduce the Repository Planning Graph (RPG), a structured representation that encodes capabilities, file structures, data flows, and functions in a unified graph. By replacing free-form natural language with an explicit blueprint, RPG enables consistent long-horizon planning for repository generation. Building on RPG, we develop ZeroRepo, a graph-driven framework that operates in three stages: proposal-level planning, implementation-level construction, and graph-guided code generation with test validation. To evaluate, we construct RepoCraft, a benchmark of six real-world projects with 1,052 tasks. On RepoCraft, ZeroRepo produces nearly 36K Code Lines and 445K Code Tokens, on average 3.9× larger than the strongest baseline (Claude Code), and 68× larger than other baselines. It achieves 81.5% coverage and 69.7% test accuracy, improving over Claude Code by 27.3 and 35.8 points. Further analysis shows that RPG models complex dependencies, enables more sophisticated planning through near-linear scaling, and improves agent understanding of repositories, thus accelerating localization. Our data and code are available at https://github.com/microsoft/RPG-ZeroRepo.