Let Agents Design How They Search: Automatic Interface Optimization for Agentic Search

A new arXiv paper (2610.29250v1) introduces SearchWeaver, a framework that automatically optimizes the interface a language-model agent uses to search a corpus, guided by downstream search performance rather than hand-tuned designs. With GPT-5.4-nano, SearchWeaver's optimized interfaces raised answer accuracy from 71.6% to 78.4% on BrowseComp-Plus and average nDCG@10 from 45.7 to 55.0 across four BRIGHT subsets, compared with Direct Corpus Interaction. The authors state the results show the value of jointly optimizing corpus artifacts, executable tools, and usage instructions to build effective search interfaces.

arXiv:2610.29250v1 Announce Type: new Abstract: Agentic search depends on the interface through which a language-model agent accesses and explores a corpus. Existing interface designs are fixed by hand and applied unchanged to every corpus, although effective designs may depend on the corpus. We introduce SearchWeaver, a framework for automatic search interface optimization guided by downstream search performance. A proposer uses execution trajectories and task rewards to jointly revise corpus artifacts, executable tools, and usage instructions. This pro cess optimizes what information is prepared in advance, which operations tools perform, and which decisions remain with the search agent. With GPT-5.4-nano, the optimized interfaces improve answer accuracy from 71.6% to 78.4% on BrowseComp-Plus and average nDCG@10 from 45.7 to 55.0 across four BRIGHT subsets over Direct Corpus Interaction. These findings highlight the value of jointly optimizing corpus artifacts, executable tools, and usage instructions to develop effective search interfaces.