Saved in:
Bibliographic Details
Main Authors: Xiu, Zidi, Sun, David Q., Cheng, Kevin, Patel, Maitrik, Date, Josh, Zhang, Yizhe, Lu, Jiarui, Attia, Omar, Vemulapalli, Raviteja, Tuzel, Oncel, Cao, Meng, Bengio, Samy
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.01357
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Next-generation AI must manage vast personal data, diverse tools, and multi-step reasoning, yet most benchmarks remain context-free and single-turn. We present ASTRA-bench (Assistant Skills in Tool-use, Reasoning \& Action-planning), a benchmark that uniquely unifies time-evolving personal context with an interactive toolbox and complex user intents. Our event-driven pipeline generates 2,413 scenarios across four protagonists, grounded in longitudinal life events and annotated by referential, functional, and informational complexity. Evaluation of state-of-the-art models (e.g., Claude-4.5-Opus, DeepSeek-V3.2) reveals significant performance degradation under high-complexity conditions, with argument generation emerging as the primary bottleneck. These findings expose critical limitations in current agents' ability to ground reasoning within messy personal context and orchestrate reliable multi-step plans. We release ASTRA-bench with a full execution environment and evaluation scripts to provide a diagnostic testbed for developing truly context-aware AI assistants.