Memory Plus Sync

Key points to consider before installing: - Provenance: the registry entry has no homepage and source is unknown. Prefer code from a verifiable repository or a trusted publisher. The README references a GitHub repo and an author email; verify that repository and author before trusting the package. - Local data access: this skill intentionally reads local chat exports (feishu/wechat/telegram) and voice logs and writes them into ~/.openclaw/memory/main.sqlite and logs under ~/.openclaw/workspace/logs. If you have private or regulated data in your chats, do not enable it without auditing and consent. - Installation ambiguity: skill.yaml lists a venv and dependencies but the registry provides no install spec. If you install, run the code in an isolated environment (create the venv yourself, inspect requirements, run tests) rather than using an automatic 'enable' flow. - Review code: the source appears to contain no network exfiltration (no HTTP/post sockets), but you should still audit the code (collector.py, memory_plus.py, monitor.py) for any changes before granting it persistent or autonomous use. - Least privilege: run first in a sandbox or on a non-sensitive account. Limit filesystem permissions for the process (e.g., run in a container or with a dedicated user) and verify file paths the skill will touch. Consider disabling autonomous invocation until you are comfortable with behavior. - Operational: if you decide to proceed, ensure backups of your memory DB, and adjust configuration to only include channels you want synced. Consider removing/stubbing the automatic collectors for channels you do not use (wechat/telegram) or adding explicit per-channel opt-in.