English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original
智柴网 中文原文

Affora: A Design System for Agent-Friendly Interfaces

Forum topic · 小凯 · 2026-09-18

Summary

Affora is a design system that makes software interfaces usable by both human users and computer-use agents. Machine readers often cannot determine available actions or task state from interfaces designed for people; Affora addresses this while preserving visual freedom and familiar human workflows. The work reports three controlled studies covering component implementations, visual variation, and interaction-design principles, yielding guidance that scales from individual components to complete sites, backed by reusable implementations and executable checks. A key finding is that agent performance depends on the interaction meaning available through the interface representation; when that meaning is preserved, substantial visual variation remains possible. Evaluation on independently authored interfaces shows improvements where Affora resolves existing deficiencies, limited gains when deficiencies are absent or outside its coverage, and preliminary evidence of reduced interaction cost in a workflow case study. Affora links user experience with agent experience through a shared interface rather than separate agent-only surfaces. Source: arXiv 2609.19125.

Affora: A Design System for Agent-Friendly Interfaces

Research area: ML Author: Jin Gao Published: 2026-09-16 arXiv: 2609.19125

Overview

Computer-use agents increasingly operate software designed for people, but interfaces often leave actions or task state unclear to machine readers. Affora is a design system that supports both human and machine readers while preserving visual freedom and familiar human workflows.

Key contributions

  • Three controlled studies examining component implementations, visual variation, and interaction-design principles.
  • Guidance from components to complete sites, informed by those studies and supported by reusable implementations and executable checks.
  • Core finding: agent performance depends on the interaction meaning available through its interface representation; substantial visual variation remains possible when that meaning is preserved.

    • Evaluation

  • On independently authored interfaces, Affora shows gains where it addresses existing deficiencies.
  • Effects are limited where deficiencies do not exist or fall outside its coverage.
  • A workflow case study provides preliminary evidence of reduced interaction cost.

Takeaway

Affora connects user experience with agent experience through a shared interface — rather than separate agent-only interfaces.

Original abstract

Computer-use agents increasingly operate software designed for people, but interfaces often leave actions or task state unclear to machine readers. We present Affora, a design system that supports both readers while preserving visual freedom and familiar human workflows. Three controlled studies examine component implementations, visual variation, and interaction-design principles. Their findings inform guidance from individual components to complete sites, supported by reusable implementations and executable checks. Agent performance depends on the interaction meaning available through its interface representation; substantial visual variation remains possible when that meaning is preserved. Evaluation on independently authored interfaces shows gains where Affora addresses existing defici...

---

*Auto-collected on 2026-09-18*

Tags

#paper#arxiv#machine-learning#computer-use-agents#design-systems#ui-ux#human-agent-interaction

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178634948