Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note3 min read

Repo0: Design-Driven Zero-to-All Code Generation

Source

arXiv

arxiv.org/abs/2608.19854v1

Read the source

Source

  • Source name: arXiv
  • Source kind: paper
  • Authors / organization: Silin Chen, Haoyi Teng, Xiaodong Gu, Yuling Shi, Jiale Huang, Yongpan Wang, Hongyu Zhang, Haibing Guan
  • Venue: arXiv: cs.SE
  • Published: 2026-08-20
  • Collected: 2026-08-21
  • URL: http://arxiv.org/abs/2608.19854v1
  • PDF: https://arxiv.org/pdf/2608.19854v1
  • DOI: UNKNOWN
  • Registry group: academic_apis_and_preprints
  • Evidence role: canonical_academic_index
  • Language: en
  • Country / region: UNKNOWN

Plain-Language Brief

The paper introduces Repo0, a framework for zero-to-all code generation that maintains an explicit architectural state using a Dual-Directed Acyclic Graph (Dual-DAG). It iteratively evolves component boundaries guided by modularity metrics and achieves high Functionality Coverage and Pass Rate in evaluations compared to existing repository-planning baselines.

What the Document Says

  • Repo0 is a framework for zero-to-all code generation that maintains an architectural state using a Dual-Directed Acyclic Graph (Dual-DAG).
  • It evolves component boundaries iteratively through structural actions guided by modularity metrics until structural convergence.
  • The converged architecture guides test-driven development code generation.
  • Evaluations were performed on six real-world repositories from RepoCraft using GPT-5 mini and DeepSeek V3.2.
  • Repo0 achieved the highest Functionality Coverage and Pass Rate across all settings.
  • Compared to RPG, the strongest repository-planning baseline, Repo0 improved Functionality Coverage by up to 20.08 percentage points and Pass Rate by up to 29.74 percentage points.
  • Ablation and structural-evolution analyses were conducted to demonstrate the importance of the Dual-DAG architectural state, modularity-guided structural evolution, and explicit structural convergence.

Key Concepts Mentioned

  • Repo0
  • zero-to-all code generation
  • Dual-Directed Acyclic Graph (Dual-DAG)
  • modularity metrics
  • functionality coverage
  • pass rate

Claims Made by the Document

  • Repo0 achieves the highest Functionality Coverage and Pass Rate compared to existing repository-planning baselines.
  • Compared to RPG, Repo0 improves Functionality Coverage by up to 20.08 percentage points and Pass Rate by up to 29.74 percentage points.

Evidence Used

  • Evaluations on six real-world repositories from RepoCraft.
  • Results from GPT-5 mini and DeepSeek V3.2.

Methods / Approach

  • iterative evolution of component boundaries
  • test-driven development code generation
  • modularity-guided structural evolution
  • structural actions

Limitations Stated by the Authors

  • UNKNOWN: The paper does not state any explicit limitations according to the provided text.

Limitations Noted by the Agent

Agent interpretation — not from the source.

  • The evaluation was conducted on six specific repositories from RepoCraft, which might limit the generalizability of the results.
  • The comparison is made with RPG, one specific repository-planning baseline. There may be other baselines that could provide different comparative results.

Open Questions

  • How does Repo0 perform on a wider range of repositories and with different repository-planning baselines?
  • What are the potential implications for industrial-scale software development?

Notes for Future Q&A

May include interpretation. Labeled where it does.

  • The paper presents a new framework for zero-to-all code generation, which is an approach that constructs an entire software project from natural-language requirements while maintaining a modular repository architecture throughout development.

Raw Abstract / Excerpt

Large language model agents have made substantial progress in code generation, yet most existing systems assume a predefined repository architecture. This assumption does not hold in zero-to-all code generation, where an agent must construct an entire software project directly from natural-language requirements while maintaining a modular repository architecture throughout development. We present Repo0, a continuous structural evolution framework for zero-to-all code generation. Repo0 maintains an explicit architectural state instantiated as a Dual-Directed-Acyclic-Graph (Dual-DAG), consisting of a requirement-level DAG, a component-level DAG, and their alignment relation. Starting from natural-language requirements, it iteratively evolves component boundaries through structural actions guided by modularity metrics until structural convergence, after which the converged architecture guides test-driven development code generation. We evaluate Repo0 on six real-world repositories from RepoCraft using GPT-5 mini and DeepSeek V3.2. Repo0 achieves the highest Functionality Coverage and Pass Rate across all settings. Compared with RPG, the strongest repository-planning baseline, Repo0 improves Functionality Coverage by up to 20.08 percentage points and Pass Rate by up to 29.74 percentage points. Ablation and structural-evolution analyses further demonstrate the importance of the Dual-DAG architectural state, modularity-guided structural evolution, and explicit structural convergence.