Search papers, labs, and topics across Lattice.
This paper investigates the economic implications of skill rewriting in large language model agents, emphasizing the trade-offs between quality and cost. By profiling skill structures and employing information-preservation strategies, the authors demonstrate that different rewriting techniques yield varying benefits across task families, with no single approach being universally superior. The proposed policy achieves a 7.0% reduction in total costs and maintains verifier quality, highlighting the importance of cost-aware skill design in operational knowledge engineering.
Skill rewriting can reduce operational costs by over 14% without sacrificing performance, challenging the notion that shorter skills are always better.
Large language model agents increasingly rely on skills: reusable procedural documents encoding workflows, tool use, implementation patterns, validation checks, and domain rules. Skill rewriting is often treated as prompt compression, but shorter skills can make agents more expensive by removing sparse operational anchors that prevent exploration, debugging, and recovery. We study skill rewriting through this economic lens. Our controlled framework profiles skill structure, rewrites skills using information-preservation strategies, and evaluates the rewrites under fixed task instructions, environments, and verifiers. Experiments on SkillsBench reveal distinct quality--cost trade-offs across strategies: API/code anchoring, workflow guarding, and rule/formula anchoring benefit different task families, with no universally dominant template. In the main held-out evaluation, the learned policy reduces total cost by 7.0\% and downstream agent-token cost by 6.0\%; in frozen cross-model transfer, the corresponding reductions average 14.7\% and 13.7\%, while verifier quality is preserved. These results position skill design as cost-aware operational knowledge engineering rather than prompt compression. Resources: \href{https://github.com/1Reminding/Skill_EE}{SkillEE}.
