GraSP: Graph-Structured Skill Compositions for LLM Agents

TL;DR

GraSP introduces a novel executable skill graph architecture that enhances LLM agent performance by structurally orchestrating skills with causal dependencies, outperforming existing methods across multiple benchmarks.

Contribution

It is the first to implement an executable skill graph with a compilation layer, transforming flat skills into DAGs for improved orchestration and robustness.

Findings

GraSP outperforms baseline methods in reward and efficiency across four benchmarks.

Performance gains increase with task complexity and are robust to skill retrieval issues.

Structured skill orchestration is more effective than simply increasing skill quantity.

Abstract

Skill ecosystems for LLM agents have matured rapidly, yet recent benchmarks show that providing agents with more skills does not monotonically improve performance -- focused sets of 2-3 skills outperform comprehensive documentation, and excessive skills actually hurt. The bottleneck has shifted from skill availability to skill orchestration: agents need not more skills, but a structural mechanism to select, compose, and execute them with explicit causal dependencies. We propose GraSP, the first executable skill graph architecture that introduces a compilation layer between skill retrieval and execution. GraSP transforms flat skill sets into typed directed acyclic graphs (DAGs) with precondition-effect edges, executes them with node-level verification, and performs locality-bounded repair through five typed operators -- reducing replanning from O(N) to O(d^h). Across ALFWorld, ScienceWorld, WebShop, and InterCode with eight LLM backbones, GraSP outperforms ReAct, Reflexion, ExpeL, and flat skill baselines in every configuration, improving reward by up to +19 points over the strongest baseline while cutting environment steps by up to 41%. GraSP's advantage grows with task complexity and is robust to both skill over-retrieval and quality degradation, confirming that structured orchestration -- not larger skill libraries -- is the key to reliable agent execution.