VibeLifeBench: Can Your Life Agent Be Proactive and Persistent in a Living World?

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Xiaohongshu Inc
• Venue: arXiv: cs.CL
• Published: 2026-08-11
• Collected: 2026-08-12
• URL: http://arxiv.org/abs/2608.10875v1
• PDF: https://arxiv.org/pdf/2608.10875v1
• DOI: UNKNOWN
• Registry group: academic_apis_and_preprints
• Evidence role: canonical_academic_index
• Language: en
• Country / region: UNKNOWN

Plain-Language Brief

Researchers introduce a new benchmark called VibeLifeBench to evaluate large language model (LLM) agents' ability to be proactive and consistent in long-horizon tasks across ten everyday life domains. The benchmark consists of 200 multi-week timelines in a simulated world with 22 mock services, where the world advances on its own and many changes are silent. The agents are graded based on their end state, timeliness of actions, and compliance with implicit constraints.

What the Document Says

• The text discusses a new benchmark called VibeLifeBench for evaluating LLM agents' performance in long-horizon tasks across ten everyday life domains.
• VibeLifeBench consists of 200 multi-week timelines in a simulated world with 22 mock services.
• The world advances on its own clock, and many changes are silent.
• Agents' performance is graded based on their end state, timeliness of actions, and compliance with implicit constraints.
• Seven frontier models were evaluated using VibeLifeBench and all scored low, indicating a gap between current agents and real-life assistance.

Key Concepts Mentioned

• large language model (LLM) agents
• VibeLifeBench
• long-horizon tasks
• everyday life domains
• multi-week timelines
• simulated world
• 22 mock services
• implicit constraints

Claims Made by the Document

• Current benchmarks for LLM agents are inadequate for everyday life assistance.
• VibeLifeBench is a new benchmark that addresses this gap.
• The seven frontier models evaluated scored low on VibeLifeBench.

Evidence Used

• The text states that existing evaluations are limited to short, self-contained requests in static environments, while everyday life assistance requires proactive and consistent agents.
• The introduction of VibeLifeBench is described as a solution for this problem.
• The evaluation results of the seven frontier models are presented as evidence of their low performance on VibeLifeBench.

Methods / Approach

• Evaluating LLM agents using VibeLifeBench, which consists of multi-week timelines in a simulated world with 22 mock services

Limitations Stated by the Authors

The text does not mention any limitations as stated by the authors.

Limitations Noted by the Agent

Agent interpretation — not from the source.

The agent's ability to fully understand or evaluate the methods and limitations is limited due to being an AI model. It cannot interpret author intent, nuanced language, or context outside of the provided text.

Open Questions

• What are the specific details of the seven frontier models that were evaluated?
• How does VibeLifeBench compare to other existing benchmarks in terms of evaluating LLM agents' performance?

Notes for Future Q&A

May include interpretation. Labeled where it does.

The text provides some context but does not offer detailed information about the specifics of the seven frontier models or a comparison between VibeLifeBench and other benchmarks.

Raw Abstract / Excerpt

Large language model (LLM) agents are increasingly deployed as personal assistants. Existing evaluations, however, mostly use short, self-contained requests in static environments. Everyday life assistance is different. A task runs for weeks rather than minutes. The world keeps changing while the agent is not being prompted. Many constraints are never stated outright. An agent that merely answers the request in front of it will fail at such a task. What is needed instead is an agent that stays proactive and consistent. It decides on its own when to act, when to ask, and when to stay silent. It notices changes that nobody announced. It keeps one plan coherent from the first day to the last. No current benchmark measures this. We introduce VibeLifeBench, a benchmark of 200 long-horizon tasks across ten everyday-life domains. Each task is a scripted multi-week timeline in a simulated world of 22 mock services. The world advances on its own clock, and many of its changes are silent, so only an agent that re-inspects the world discovers them. Every task is graded by fine-grained, weighted checks that read only what the agent actually left behind, covering the end state, the timeliness of its actions, and whether it upheld the implicit constraints. We evaluate seven frontier models. All of them score low, which shows how far current agents are from assisting with real life. We will open-source all tasks, environments, and the evaluation framework.