Scaling Coding Agents via Atomic Skills

• URL: http://arxiv.org/abs/2604.05013v1
• Date: Mon, 06 Apr 2026 16:36:21 GMT
• Title: Scaling Coding Agents via Atomic Skills
• Abstract summary: We first formalize five fundamental atomic skills, code localization, code editing, unit-test generation, issue reproduction, and code review.<n>Compared with composite coding tasks, these atomic skills are more general and composable.<n>In this manner, atomic skills are consistently improved without negative interference or trade-offs between them.
• Score: 23.614587850545988
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Current LLM coding agents are predominantly trained on composite benchmarks (e.g., bug fixing), which often leads to task-specific overfitting and limited generalization. To address this, we propose a novel scaling paradigm that shifts the focus from task-level optimization to atomic skill mastery. We first formalize five fundamental atomic skills, code localization, code editing, unit-test generation, issue reproduction, and code review, that serve as the basis vectors for complex software engineering tasks. Compared with composite coding tasks, these atomic skills are more generalizable and composable. Then, we scale coding agents by performing joint RL over atomic skills. In this manner, atomic skills are consistently improved without negative interference or trade-offs between them. Notably, we observe that improvements in these atomic skills generalize well to other unseen composite coding tasks, such as bug-fixing, code refactoring, machine learning engineering, and code security. The observation motivates a new scaling paradigm for coding agents by training with atomic skills. Extensive experiments demonstrate the effectiveness of our proposed paradigm. Notably, our joint RL improves average performance by 18.7% on 5 atomic skills and 5 composite tasks.

Related papers

• Dual-Process Atomic Skill Learning: Decoupling Semantic Reasoning and Real-Time Control [21.97625864571452]
  Language-conditioned Imitation Learning (IL) is essential for enabling robots to perform complex tasks following natural language instructions.<n>We propose Dual-Process Atomic Skill Learning (DASL), a novel hierarchical imitation learning framework that decouples slow semantic reasoning from real-time motion control.<n>DASL significantly outperforms state-of-the-art baselines, excelling in skill acquisition and compositional generalization to unseen instructions.
  arXiv  Detail & Related papers  (2026-07-12T07:38:53Z)
• ATOM-Bench: A Real-World Benchmark for Atomic Skills and Compositional Generalization in Manipulation Policies [18.55273532676497]
  Generalist manipulation policies are increasingly presented as foundation models for robotic control.<n>ATOM-Bench is a real-world benchmark for evaluating both atomic skills and compositional generalization in manipulation policies.
  arXiv  Detail & Related papers  (2026-06-15T15:08:42Z)
• SkillComposer: Learning to Evolve Agent Skills for Specification and Generalization [99.18305770138677]
  Agent skills are reusable strategies that guide agent reasoning and action.<n>Current skill construction methods treat the problem as one-shot extraction.<n>We introduce SkillComposer, a framework that decomposes skill construction into three learnable operations.
  arXiv  Detail & Related papers  (2026-06-04T12:16:08Z)
• Agent psychometrics: Task-level performance prediction in agentic coding benchmarks [24.348135523715815]
  We present a framework for predicting success or failure on individual tasks tailored to the agentic coding regime.<n>Our approach augments Item Response Theory (IRT) with rich features extracted from tasks, including issue statements, repository contexts, solutions, and test cases.
  arXiv  Detail & Related papers  (2026-04-01T07:59:59Z)
• AtomicVLA: Unlocking the Potential of Atomic Skill Learning in Robots [92.18094199070693]
  Real-world robotic tasks often involve long-horizon, multi-step problem-solving and require generalization for continual skill acquisition.<n>We propose AtomicVLA, a unified planning-and-execution framework that jointly generates task-level plans, atomic skill abstractions, and fine-grained actions.<n>In simulation, AtomicVLA outperforms $_0$ by 2.4% on LIBERO, 10% on LIBERO-LONG, and outperforms $_0$ and $_0.5$ by 0.22 and 0.25 in average task length on CALVIN
  arXiv  Detail & Related papers  (2026-03-08T14:18:56Z)
• SkillCraft: Can LLM Agents Learn to Use Tools Skillfully? [67.69996753743129]
  We introduce SkillCraft, a benchmark explicitly stress-test agent ability to form and reuse higher-level tool compositions.<n> SkillCraft features realistic, highly compositional tool-use scenarios with difficulty scaled along both quantitative and structural dimensions.<n>We propose a lightweight evaluation protocol that enables agents to auto-compose atomic tools into executable Skills, cache and reuse them inside and across tasks.
  arXiv  Detail & Related papers  (2026-02-28T15:44:31Z)
• Hybrid-Gym: Training Coding Agents to Generalize Across Tasks [59.95803522351185]
  In this paper, we describe some transferable skills that are shared across diverse tasks.<n>We propose a training environment, Hybrid-Gym, consisting of a set of scalable synthetic tasks.<n>Experiments show that agents trained on our synthetic tasks effectively generalize diverse real-world tasks.
  arXiv  Detail & Related papers  (2026-02-18T19:30:55Z)
• SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning [83.98129545309277]
  We propose SkillRL, a framework that bridges the gap between raw experience and policy improvement.<n>Our approach introduces an experience-based distillation mechanism to build a hierarchical skill library SkillBank.<n> Experimental results on ALF, WebShop and seven search-augmented tasks demonstrate that SkillRL achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2026-02-09T03:17:17Z)
• Cogito, ergo sum: A Neurobiologically-Inspired Cognition-Memory-Growth System for Code Generation [9.920563105290894]
  Cogito is a neurobiologically inspired multi-agent framework to enhance the problem-solving capabilities in code generation tasks with lower cost.<n>Cogito accumulates knowledge and cognitive skills at each stage,ultimately forming a Super Role an all capable agent to perform the code generation task.
  arXiv  Detail & Related papers  (2025-01-30T01:41:44Z)
• AtomThink: Multimodal Slow Thinking with Atomic Step Reasoning [68.65389926175506]
  We propose a novel paradigm of Self-structured Chain of Thought (SCoT)<n>Our method can not only generate cognitive CoT structures for various complex tasks but also mitigates the phenomena of overthinking for easier tasks.<n>We conduct extensive experiments to show that the proposed AtomThink significantly improves the performance of baseline MLLMs.
  arXiv  Detail & Related papers  (2024-11-18T11:54:58Z)
• Laying the Foundation First? Investigating the Generalization from Atomic Skills to Complex Reasoning Tasks [40.7766635942194]
  We propose a probing framework to investigate whether the atomic skill can spontaneously generalize to complex reasoning tasks. We then introduce a hierarchical curriculum learning training strategy to achieve better skill generalization. By leveraging hierarchical curriculum learning, we successfully induce generalization, significantly improve the performance of open-source LMs on complex reasoning tasks.
  arXiv  Detail & Related papers  (2024-03-14T15:20:54Z)
• RL-GPT: Integrating Reinforcement Learning and Code-as-policy [82.1804241891039]
  We introduce a two-level hierarchical framework, RL-GPT, comprising a slow agent and a fast agent. The slow agent analyzes actions suitable for coding, while the fast agent executes coding tasks. This decomposition effectively focuses each agent on specific tasks, proving highly efficient within our pipeline.
  arXiv  Detail & Related papers  (2024-02-29T16:07:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.