← 返回 Skills 市场
randomradio

Memoria

作者 i.an · GitHub ↗ · v0.2.0 · MIT-0
cross-platform ⚠ suspicious
105
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install thememoria
功能描述
Use Memoria as OpenClaw's durable memory slot. Triggers: "remember this", "save to memory", "what do you remember", "continue from last time", "forget this",...
安全使用建议
This skill appears to do what it says (use Memoria as durable memory) but pay attention before installing: 1) The SKILL.md/setup docs ask you to supply MEMORIA_API_URL/API_KEY and an embedding key — those secrets are required for cloud/embedded operation but were not declared in the registry metadata; confirm where those keys will be stored and which network endpoint you will trust. 2) The setup doc suggests running a remote install script via curl | bash; avoid piping unknown scripts into your shell unless you audit the script or use a vetted release. 3) Prefer installing via the official openclaw plugin registry or a checked GitHub release and verify signatures or pinned commit SHAs where possible. 4) Review the upstream Memoria repo (https://github.com/matrixorigin/Memoria) and its install scripts before running them. 5) If you will store sensitive data, ensure the Memoria backend and API endpoint meet your security and privacy requirements. If you want higher confidence, ask the publisher to: add declared env vars to the registry metadata, include a vetted install spec, and avoid recommending curl|bash in their user-facing docs.
功能分析
Type: OpenClaw Skill Name: thememoria Version: 0.2.0 The skill bundle provides instructions for integrating Memoria as a long-term memory system for OpenClaw. It is classified as suspicious because `references/setup.md` contains a 'curl | bash' command (targeting `https://raw.githubusercontent.com/matrixorigin/Memoria/main/scripts/install.sh`) for local installation, which is a high-risk pattern for remote code execution. While the instructions appear aligned with the stated purpose of the plugin and lack clear malicious intent, this installation method represents a significant security risk if executed by the agent.
能力评估
Purpose & Capability
The name, description, and runtime instructions consistently describe using Memoria as OpenClaw's durable memory and map to the listed memory_* operations. However, the SKILL.md and setup docs reference API_URL/API_KEY/EMBEDDING_API_KEY and explicit plugin installation steps that are not reflected in the registry's declared requirements (registry lists no required env vars or binaries). This mismatch between declared requirements and the documented onboarding is unexpected.
Instruction Scope
The runtime instructions focus on retrieving, storing, correcting, and managing memory via memory_* tools and are scoped to the stated purpose. The setup instructions (in references/setup.md) include network operations (git clone, and a curl | bash installer) and request API keys for cloud/local backends; these are related to enabling the plugin but broaden the surface the operator must trust.
Install Mechanism
The registry has no formal install spec, but the included setup docs recommend: (a) openclaw plugin install (safe/expected) and (b) optionally cloning from GitHub or running curl -sSL https://raw.githubusercontent.com/.../install.sh | bash. Piping a remote install script to bash is a high-risk install mechanism because it executes unreviewed code fetched at runtime. The git clone fallback is less risky but still pulls code from an external repo. The instructions do not provide cryptographic verification or pinned release URLs.
Credentials
The skill registry declares no required environment variables or primary credential, but the SKILL.md and setup docs explicitly reference MEMORIA_API_URL, API_KEY, and EMBEDDING_API_KEY for cloud and embedded modes. Asking for API keys and endpoints is proportionate to a memory plugin, but the mismatch (documentation requires secrets while registry metadata lists none) is a coherence problem and a user-config/expectation risk.
Persistence & Privilege
The skill does not request always:true and does not declare modifications to other skills or system-wide settings. It is user-invocable and allows autonomous invocation (the platform default). Nothing in the skill requests elevated persistence beyond normal plugin behavior.
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 thememoria
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /thememoria 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
v0.2.0
Initial ClawHub release of the Memoria durable-memory skill for OpenClaw.
元数据
Slug thememoria
版本 0.2.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Memoria 是什么?

Use Memoria as OpenClaw's durable memory slot. Triggers: "remember this", "save to memory", "what do you remember", "continue from last time", "forget this",... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 105 次。

如何安装 Memoria?

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

Memoria 是免费的吗?

是的,Memoria 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Memoria 支持哪些平台?

Memoria 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 Memoria?

由 i.an(@randomradio)开发并维护,当前版本 v0.2.0。

💬 留言讨论