help stay focus . currently only support chinese
WarnAudited by ClawScan on May 10, 2026.
Overview
This focus-monitoring skill is purpose-aligned, but it asks to run unreviewed remote code and to continuously capture, store, and upload screen images, so it needs careful review before use.
Install only if you are comfortable with whole-screen monitoring. Before running it, inspect the GitHub repository and start.bat, confirm which vision model receives screenshots, decide what data should be excluded, and make sure you know how to stop monitoring and delete stored FocusOS_Data history.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Installing it could run unreviewed code on your computer and start a local monitoring service.
The skill is instruction-only, but it directs the agent to fetch and run a mutable remote repository and batch file that were not included in the reviewed artifacts.
git clone https://github.com/HR2AY/focusAI ... cd focusAI && start.bat
Review the GitHub repository and start.bat yourself, pin a trusted commit or checksum, and only run it after explicit approval in a constrained environment.
Private information visible on your screen could be stored, analyzed later, or sent to an external vision model.
The skill periodically captures screenshots, uploads them for model analysis, stores activity history locally, and reuses that history for later retrieval; the artifacts do not specify clear retention, exclusion, deletion, or provider boundaries.
privacy_note: 会定期截屏并上传至视觉模型分析 ... 数据默认位于 `~/Desktop/FocusOS_Data/` ... 长期记忆检索 (RAG 架构)
Use only when needed, confirm the exact model/provider, define excluded apps or content, set retention and deletion rules, and review the stored FocusOS_Data files regularly.
The agent could start screen monitoring or change how often screenshots are taken if the local service is running.
The local API gives the agent control to start monitoring and change screenshot/model settings. This is aligned with the skill’s purpose, but it is sensitive and should require explicit user confirmation.
POST `/start` | 启动监控 ... POST `/config` | 覆写配置（切换模型或截图间隔）
Require clear user confirmation before calling start or config endpoints, especially when changing model providers or screenshot intervals.
The monitoring process may keep running and collecting screen data until you stop or close it.
The skill describes background monitoring and a GUI that may stay on top. This is disclosed and purpose-aligned, but it is persistent behavior the user should consciously control.
后台监控 | 必须告知用户截屏行为、内存和 API 使用情况 ... 前端 GUI | 启动前需询问用户，告知 UI 会置顶，需在任务栏关闭
Verify how to stop monitoring, close the GUI, and remove stored data before using it for sensitive work.