Web agents reuse skills to retrieve live economic data more efficiently
EconSkills: Studying Skill Transfer and Retrieval for Web Agents on Live Economic Data
Artificial IntelligenceComputation and Language
Summary
Web agents often revisit the same websites to gather economic data, but usually don’t reuse what they learned before. The authors created EconSkills, a system that saves and organizes these learning steps into reusable procedures or skills. They tested how well these skills transfer to new but related tasks, showing that using abstracted procedural skills helps agents succeed more often and faster than just replaying past actions. This approach helps web agents navigate economic sites more efficiently when gathering live data.
What this means in practice
- •For financial data teams: Improve automation of economic data collection from repeated web portals by reusing verified navigation procedures to reduce errors and speed up retrieval.
- •For web automation developers: Build smarter web agents that adapt and select reusable skills for retrieving specific live data across multiple economic websites.
Authors
Yinzhu Quan, Zefang Liu
Abstract
Web agents often revisit the same sites, yet most evaluations discard the procedures learned in earlier successful interactions. We introduce EconSkills, a skill library and evaluation framework that distills verified EconWebArena trajectories into parameterized standard operating procedures for retrieving live economic data. Each skill records its scope, navigation procedure, site-specific guidance, verification checks, and recovery steps while replacing source-instance values with placeholders. EconSkills separates two questions: whether a known relevant procedure transfers to a held-out task, and whether an agent can retain that benefit when selecting from a library. In controlled transfer, matched skills improve success over no-skill prompting and require fewer steps on paired successes, while abstraction is substantially more effective than replaying raw trajectories. At library scale, retrieval is competitive with the no-skill baseline overall and performs best on directly covered tasks; coverage-stratified outcomes show that approximate matches on uncovered tasks offset these gains. Browser trajectories further identify when procedural guidance shortens portal-specific navigation and when semantic verification remains necessary. These results establish that reusable economic web procedures can transfer across task instances and provide a concrete design target for coverage-aware selection and context delivery.