SkillForge: Evolving Verifiable Skills for Reinforcement Learning Agents

2026-08-25

AI Summary

为解决RL训练的LLM代理无法跨情节积累可重用知识的问题，提出SkillForge框架，通过环境互动验证和优化技能，提升代理性能。

Abstract

Large language model (LLM) agents are trained with reinforcement learning (RL) for complex decision-making tasks. However, most RL-trained agents remain episodic and cannot accumulate reusable knowledge across episodes. Recent skill-based approaches, such as SkillRL, attempt to address this issue by extracting skills from raw trajectories, but treat the skill bank as an append-only repository without verifying whether stored skills remain effective. In this paper, we propose SkillForge, a framework for continuous skill evolution that enables skills to be verified and refined through environment interaction. By making skill usage explicit during agent interaction, RL can directly optimize both environment actions and skill invocation decisions. SkillForge further introduces evidence-based skill verification and multi-pathway skill induction, allowing the skill bank to continuously grow while maintaining its quality. Extensive experiments on ALFWorld, WebShop, and AppWorld show that SkillForge consistently outperforms SkillRL, demonstrating the effectiveness of continuously verified skills in training stronger LLM agents.

Problem

• Reinforcement Learning
• Skill Evolution
• Knowledge Accumulation
• Skill Verification

Innovation

• Skill Evolution
• Evidence-based Verification
• Multi-pathway Induction