Claw-Anything: Benchmarking Always-On Personal Assistants with Broader Access to User's Digital World
2026-05-25 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors created a new test called Claw-Anything to better measure how well AI assistants can understand and help with a wide range of user activities over time. Unlike older tests, this one includes long-term user histories, connected services, and interactions on different devices, making it more realistic and challenging. They found that even advanced AI like GPT-5.5 struggles with these tasks, showing current AI isn't yet great at being always-ready helpers. The authors also built a system to automatically create lots of training data, which helped improve AI performance.
Large Language ModelsBenchmarksContextual ReasoningPersonal AssistantsUser Activity HistoryBackend ServicesGUI and CLIMulti-device InteractionProactive AssistanceAutomated Data Generation
Authors
Yusong Lin, Xinyuan Liang, Haiyang Wang, Qipeng Gu, Siqi Cheng, Jiangui Chen, Shuzhe Wu, Feiyang Pan, Lue Fan, Sanyuan Zhao, Dandan Tu
Abstract
Large language model agents are increasingly envisioned as always-on personal assistants with access to anything relevant in the user's digital world. Yet current systems operate over only narrow slices of that world, limiting context-sensitive reasoning and effective assistance. Existing benchmarks similarly provide only partial user state and therefore fail to capture performance in such a broad, always-on setting. To address this gap, we introduce Claw-Anything, a benchmark that expands agent context along three dimensions: long-horizon activity histories, interdependent backend services, and integrated GUI and CLI interaction across multiple devices. To instantiate this setting, we simulate months of user activity through multi-round event injection, producing complex world states and realistic noise, including irrelevant events and conflicting signals. Agents must reason over rich contextual environments while remaining robust to such noise. This expanded scope also enables the evaluation of proactive assistance, requiring agents to anticipate user needs and deliver timely recommendations. Experiments show that GPT-5.5 achieves only 34.5% pass@1, substantially below prior benchmarks, underscoring a gap between current agent capabilities and the demands of always-on personal assistance. Alongside the benchmark, we release an automated data-generation pipeline that yields 2,000 training environments and improves the base model by 23.7%, demonstrating its utility of scalable data infrastructure.