Desktop interaction streams provide a continuous, privacy-sensitive record of interleaved user tasks. Transforming these streams into task-organized personal logs on-device faces two main challenges: exhaustive Vision-Language Model (VLM) processing strains local resources, and global stream processing causes cross-task context pollution. We present FOCAL (Filtered On-device Continuous Activity Logging), a privacy-first multi-agent system utilizing a unified filter-plan-log architecture. It cascades a lightweight Filter Agent for noise suppression, a text-only Brain Agent for task attribution, a Record Agent for selective visual reasoning, and a task-isolated Memory Agent for context-coherent summarization. Experiments on DesktopBench (comprising 2,572 screenshots across 420 complex sessions) show FOCAL reduces total token consumption by 60.4% and VLM call count by 72.3% versus a baseline, while boosting Key Information Recall (KIR) from 0.38 to 0.61. Crucially, under $A{\to}B{\to}A$ task interruptions, FOCAL maintains Task Acc 0.81 and KIR 0.80, whereas the baseline collapses to Task Acc 0.03. FOCAL pioneers the efficient, on-device summarization of instruction-free desktop streams into multi-perspective personal logs.

翻译：桌面交互流提供了交织用户任务的连续且隐私敏感的记录。将这些流转换为设备端按任务组织的个人日志面临两大挑战：穷举式视觉-语言模型（Vision-Language Model, VLM）处理会消耗本地资源，而全局流处理则会导致跨任务上下文污染。我们提出FOCAL（过滤式设备端持续活动日志记录），一个采用统一“过滤-计划-记录”架构的隐私优先多智能体系统。该系统级联了用于噪声抑制的轻量级过滤智能体（Filter Agent）、用于任务归因的纯文本大脑智能体（Brain Agent）、用于选择性视觉推理的记录智能体（Record Agent）以及用于上下文连贯摘要的任务隔离记忆智能体（Memory Agent）。在DesktopBench（包含420个复杂会话中的2,572张截图）上的实验表明，与基线相比，FOCAL将总令牌消耗降低了60.4%，VLM调用次数减少了72.3%，同时将关键信息召回率（Key Information Recall, KIR）从0.38提升至0.61。至关重要的是，在A→B→A任务中断场景下，FOCAL保持了0.81的任务准确率（Task Acc）和0.80的KIR，而基线则崩溃至Task Acc 0.03。FOCAL开创了将无指令桌面流高效设备端摘要为多视角个人日志的先河。