Voice Memo Sync

Sync, transcribe, and intelligently organize voice memos, audio/video files, and URLs. 同步、转录、智能整理语音备忘录、音视频文件和视频链接。

This skill appears to do what it says, but review these points before installing: (1) It will read local iCloud/Voice Memos directories and your agent memory files (USER.md / MEMORY.md) for personalized LLM processing — consider whether you want those read. (2) Optional components (whisper-cpp, yt-dlp, summarize, remindctl) enable extra features and may perform network access (model download from HuggingFace, downloading videos). If you prefer no network downloads, pre-install models/tools manually or inspect/modify scripts. (3) The installer can add a heartbeat auto-sync entry to HEARTBEAT.md — opt out if you don't want scheduled automatic processing. (4) Because the skill creates Apple Notes and Reminders via AppleScript, review those scripts if you need stricter controls. If these behaviors are acceptable, the skill is coherent with its purpose; otherwise, inspect/adjust the scripts and configuration before use.

Type: OpenClaw Skill Name: voice-memo-sync Version: 1.6.1 The skill performs high-privilege operations including accessing private macOS Library containers for Voice Memos and iCloud, and using 'osascript' to automate Apple Notes and Reminders. While these actions are aligned with the stated purpose of syncing and organizing memos, the scripts (specifically 'process.sh' and 'create-apple-note.sh') execute shell commands and system automation that could be abused if provided with malicious input. Additionally, 'process.sh' performs an unverified remote download of a model binary from HuggingFace via 'curl', and 'install.sh' establishes persistence by modifying the OpenClaw 'HEARTBEAT.md' file.