Zettelkasten CN

卢曼卡片学习法（Zettelkasten）全生命周期管理系统，支持九大人生领域分类、撤销操作、交互式创建流程。处理闪念笔记、文献笔记、永久笔记、项目笔记的增删改查，以及与Agent记忆系统的双向关联。

This skill is a local-file Zettelkasten manager and appears consistent with that purpose. Before installing or running it: 1) Review and, if desired, override the default storage location by setting ZETTELKASTEN_CARDS_DIR so files are written where you expect (the default is ~/Desktop/cardsdata). 2) Note the skill writes and moves files (create/delete/undo) and keeps an operation history in the .system folder; back up any important notes first. 3) If you run the interactive/quick create flows via an autonomous agent, be aware the agent will execute the bundled Python scripts locally and may create/modify files — only allow autonomous invocation if you trust the skill. 4) The docs contain a small mismatch about the history file path; inspect the bundled scripts (card_manager.py / undo_manager.py) to confirm actual behavior. If you want higher assurance, run the scripts manually in a sandboxed environment or review the code lines that set CARDS_DIR and HISTORY_FILE to confirm they point to acceptable locations.

Type: OpenClaw Skill Name: zettelkasten-cn Version: 1.0.0 The OpenClaw AgentSkills skill bundle implements a Zettelkasten note management system. It performs file system operations (create, read, update, delete, move files) within user-designated directories (`~/Desktop/cardsdata`) and the agent's workspace (`~/.openclaw/workspace/memory`). The skill uses `subprocess.run` to execute its own local Python scripts (`card_manager.py`, `undo_manager.py`) by passing arguments as a list, which prevents shell injection vulnerabilities. Filenames are sanitized, and user input is written to files as content, not executed. The `SKILL.md` and `references/advanced.md` files provide functional instructions and examples without any evidence of prompt injection attempts or malicious directives. All operations are directly aligned with the stated purpose of a note management system and are implemented with reasonable security practices, lacking any clear intent for harmful behavior like data exfiltration, backdoors, or unauthorized remote control.