视觉系幻灯片捕手 (The Slide Sniper)

后台监控全屏视频或直播，利用视觉模型检测幻灯片翻页，自动截图提取文字并排版到笔记软件中。

This skill does what it says, but it captures the screen repeatedly — which can collect private information — and it declares network access even though the instructions describe only local OCR and file writes. Before installing or enabling it: 1) Confirm whether network access will be used (cloud OCR or syncing). If you do not trust that, deny network permission. 2) Require explicit user consent each time the skill runs (avoid autonomous background runs). 3) Test it on non-sensitive content first to verify the 'privacy filtering' actually excludes chat/notifications. 4) Restrict the write directory and periodically audit or delete saved screenshots. 5) If you must use cloud OCR, insist on a documented, auditable endpoint and minimal credentials. If any of these items are unclear or you cannot limit network access, treat the skill as high-risk and avoid installing it.

Type: OpenClaw Skill Name: slide-sniper Version: 0.1.0 The skill's purpose is to monitor video streams for slide changes, screenshot, extract text, and compile notes. Permissions are appropriately scoped, explicitly disallowing shell access (`shell: allowed: false`) and restricting filesystem write access to a dedicated notes directory (`~/Documents/Notes/SlideSniper`). While network access is allowed, the skill's instructions in `SKILL.md` do not direct the agent to perform any network-related operations, data exfiltration, or unauthorized communication. The instructions also include positive security 'red lines' to prevent interference or privacy breaches. There is no evidence of malicious intent, prompt injection for harmful purposes, or exploitation of vulnerabilities.