Reusable code components improve building large software repositories

Large-scale Repository Engineering via Agent-Native Reusable Code Primitives

Software EngineeringComputation and LanguageComputer Vision and Pattern Recognition

Summary

Building large software projects is hard because many parts need to work well together, like code modules, tests, and dependencies. The paper introduces 'Code Primitives,' smaller reusable code pieces that can automatically adapt to specific projects using a language model. They created LEGO, a system that picks and adapts these code primitives to build a whole repository, checking and fixing errors along the way. Their tests show LEGO improves results over previous methods, making software construction more reliable at scale.

What this means in practice

  • For software engineering teams: Automatically assemble large codebases from reusable and adaptable code components validated for integration and testing.
  • For devops teams: Use adapted code primitives to ensure consistent integration of modules, dependencies, and tests in complex software pipelines.

Authors

Haibo Jin, Peng Kuang, Xucheng Yu, Jerry Wang, Dehao Wu, Haohan Wang

Abstract

Large language models equipped with development environments have moved code generation toward repository-scale construction, yet building complete repositories remains difficult because interacting modules, interfaces, configurations, tests, and dependencies must work together. We introduce Code Primitives, agent-native reusable executable components with interface contracts, dependency closures, validation tests, and provenance. Each primitive uses a resident LLM to assess relevance and adapt its implementation, interfaces, and dependencies to the target repository, and we organize 1,424 validated primitives in CodeFace, a searchable library for repository construction. We introduce LEGO (Large-scale repository Engineering via aGent-native reusable cOde primitives), which activates task-relevant primitives, integrates their adapted implementations with task-specific code while resolving cross-component constraints, and revises the result against executed tests. To measure construction end to end, we build LEGO-REPO, a benchmark of 522 executable reconstruction tasks spanning seven software domains, 22 capability tracks, and five difficulty levels, scored against native test suites between an empty-package floor and original-source ceiling. The strongest of 13 evaluated backbones reaches a delivery score of 0.318 and scores zero on 41.0% of tasks; LEGO improves all 13 by 0.1474 on average and raises GPT-5.6-terra from 0.3180 to 0.5134 (+61.4%). In controlled comparisons, adapted primitives outperform retrieved code supplied as context or vendored unchanged. The effect persists against independent repository agents, across three external benchmarks, and with a disjointly re-mined CodeFace; GPT-OSS-20B for adaptation and diagnosis retains 95.1% of the homogeneous score at 24.0% lower cost.