← 返回 Skills 市场
aeoleader

Memory On Demand

作者 aeoleader · GitHub ↗ · v1.0.0
cross-platform ⚠ suspicious
1622
总下载
0
收藏
12
当前安装
1
版本数
在 OpenClaw 中安装
/install memory-on-demand
功能描述
按需记忆检索。当用户询问历史相关问题时,自动搜索 memory 和 QMD 获取相关信息。
安全使用建议
This skill will search your local QMD index and files under ~/.openclaw/workspace/memory/ when trigger keywords appear. Before installing: (1) confirm you are comfortable with the agent reading those local memory files; (2) ensure the referenced tools (qmd, memory_search) exist on your system — the metadata doesn't declare them explicitly; (3) verify what 'QMD' and the workspace configuration refer to in your environment; and (4) consider whether automatic triggering on the listed keywords is acceptable or if you prefer an explicit user opt-in to avoid accidental retrieval of sensitive memories. If you want higher assurance, ask the author to declare required binaries and to provide a minimal test log showing exactly which commands will run and what files will be read.
功能分析
Type: OpenClaw Skill Name: memory-on-demand Version: 1.0.0 The `SKILL.md` file instructs the AI agent to execute shell commands like `grep -r "关键词" ~/.openclaw/workspace/memory/` and `qmd search "关键词"`. The '关键词' placeholder is intended to be filled with user input. If the agent does not rigorously sanitize this user input before embedding it into the shell commands, it creates a significant shell injection vulnerability, potentially leading to arbitrary command execution (RCE). This is a critical vulnerability, not direct malice, making the skill suspicious.
能力评估
Purpose & Capability
The skill's name/description (on-demand memory retrieval) matches the actions in SKILL.md: searching a QMD index and local memory files. Requesting access to a memory folder is expected for this purpose. One minor inconsistency: the instructions call external tools (qmd, memory_search) although the skill metadata declares no required binaries.
Instruction Scope
Instructions explicitly tell the agent to run 'qmd search' and grep the ~/.openclaw/workspace/memory/ directory. Reading those local files is within the stated purpose, but the SKILL.md gives the agent direct permission to read the user's memory directory and to invoke qmd/memory_search commands — confirm you are comfortable with that level of local data access.
Install Mechanism
There is no install spec and no code files (instruction-only), so nothing will be written to disk by an installer. This lowers risk; however, the runtime relies on external tools that must already exist on the system (qmd, optional memory_search).
Credentials
No environment variables, credentials, or config paths are requested in the metadata. The SKILL.md only references a local workspace path (~/.openclaw/workspace/memory/), which is proportionate to a memory-retrieval skill.
Persistence & Privilege
The skill is not always-on and is user-invocable; it does not request elevated or persistent platform privileges. Autonomous invocation is allowed by default, which is normal for skills and not an extra privilege here.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install memory-on-demand
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /memory-on-demand 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
- Initial release of memory-on-demand skill for automatic memory retrieval. - Detects and responds to user queries involving historical context using keywords (e.g., "之前", "历史", "还记得吗"). - Prioritizes QMD search for faster, more accurate results; uses memory file search as a fallback. - Returns organized search results to the user automatically. - Minimizes resource use by only searching memory when needed.
元数据
Slug memory-on-demand
版本 1.0.0
许可证
累计安装 13
当前安装数 12
历史版本数 1
常见问题

Memory On Demand 是什么?

按需记忆检索。当用户询问历史相关问题时,自动搜索 memory 和 QMD 获取相关信息。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1622 次。

如何安装 Memory On Demand?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install memory-on-demand」即可一键安装,无需额外配置。

Memory On Demand 是免费的吗?

是的,Memory On Demand 完全免费(开源免费),可自由下载、安装和使用。

Memory On Demand 支持哪些平台?

Memory On Demand 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Memory On Demand?

由 aeoleader(@aeoleader)开发并维护,当前版本 v1.0.0。

💬 留言讨论