SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning
This addresses the challenge of inefficient experience reuse in LLM agents for complex tasks, offering a novel framework with significant performance gains.
The paper tackles the problem of LLM agents failing to learn from past experiences due to redundant raw trajectories, proposing SkillRL for automatic skill discovery and recursive evolution, which achieves state-of-the-art performance with over 15.3% improvement on benchmarks like ALFWorld and WebShop.
Large Language Model (LLM) agents have shown stunning results in complex tasks, yet they often operate in isolation, failing to learn from past experiences. Existing memory-based methods primarily store raw trajectories, which are often redundant and noise-heavy. This prevents agents from extracting high-level, reusable behavioral patterns that are essential for generalization. In this paper, we propose SkillRL, a framework that bridges the gap between raw experience and policy improvement through automatic skill discovery and recursive evolution. Our approach introduces an experience-based distillation mechanism to build a hierarchical skill library SkillBank, an adaptive retrieval strategy for general and task-specific heuristics, and a recursive evolution mechanism that allows the skill library to co-evolve with the agent's policy during reinforcement learning. These innovations significantly reduce the token footprint while enhancing reasoning utility. Experimental results on ALFWorld, WebShop and seven search-augmented tasks demonstrate that SkillRL achieves state-of-the-art performance, outperforming strong baselines over 15.3% and maintaining robustness as task complexity increases. Code is available at this https://github.com/aiming-lab/SkillRL.