Write, Execute, Refine: From Skill Follower to Skill Optimizer via Reinforcement Learning from Execution Feedback

arXiv:2608.17587 · cs.CL · Submitted 2026-08-18 · Read on arXiv

cs.CL

Submitted: 2026-08-18

Updated: 2026-10-13

Code: https://github.com/littlepkk/WER4skill-optimizer-training

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Reinforcement Learning — 1× in this paper · explained in 4 episodes

Sources

• EvoSkill: Automated Skill Discovery for Multi-Agent Systems
• $\tau^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment
• ReSkill: Reconciling Skill Creation with Policy Optimization in Agentic RL
• SoK: Agentic Skills -- Beyond Tool Use in LLM Agents
• SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks
• Contextual Experience Replay for Self-Improvement of Language Agents
• SkillRevise: Improving LLM-Authored Agent Skills via Trace-Conditioned Skill Revision
• SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization
• Trace2Skill: Distill Trajectory-Local Lessons into Transferable Agent Skills
• SkillOS: Learning Skill Curation for Self-Evolving Agents
• Toolformer: Language Models Can Teach Themselves to Use Tools
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning
• Skill-R1: Agent Skill Evolution via Reinforcement Learning
• Voyager: An Open-Ended Embodied Agent with Large Language Models
• Reinforcement Learning for Self-Improving Agent with Skill Library
• SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning
• Qwen3 Technical Report
• SkillMaster: Toward Autonomous Skill Mastery in LLM Agents
• SkillOpt: Executive Strategy for Self-Evolving Agent Skills

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering