SkillX: Automatically Constructing Skill Knowledge Bases for Agents

Researchers introduce SkillX, an automated framework for building reusable skill knowledge bases for AI agents that addresses inefficiencies in current self-evolving paradigms. The system uses multi-level skill design, iterative refinement, and exploratory expansion to create plug-and-play skill libraries that improve task success and execution efficiency across different agents and environments.

• →SkillX creates reusable skill knowledge bases that can be shared across different AI agents and environments, reducing redundant learning.
• →The framework uses a three-tiered hierarchy of strategic plans, functional skills, and atomic skills to organize agent capabilities.
• →Iterative refinement automatically improves skill quality based on execution feedback from real-world performance.
• →Exploratory expansion proactively generates novel skills beyond initial training data to expand coverage.
• →Testing on challenging benchmarks shows consistent improvements in task success rates and execution efficiency for weaker base agents.