← Back to Skills Marketplace
OpenClaw 记忆系统
by
Nick-Liu1989
· GitHub ↗
· v1.0.0
· MIT-0
202
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install openclaw-memory-system
Description
支持多模态图片和工具调用记忆，细粒度项目和用户隔离，及基于自然语言的记忆修正与版本控制。
Usage Guidance
This package's docs and manifest describe many runtime JS modules, but the archive you provided only includes documentation and a small install script — the modules SKILL.md references (image-memory.js, tool-memory.js, memory-correct.js, tests, etc.) are missing. Before installing: 1) verify the full source (e.g., the claimed GitHub repo) contains the runtime files; ask the author for a complete release; 2) inspect any postinstall or install scripts shipped with the real package — here package.json refers to postinstall.js which is absent; 3) avoid running scripts or tests from untrusted packages until you confirm their contents; 4) if you enable Feishu/webhook integrations, use webhooks you control and review what data will be sent. The current inconsistencies look like an incomplete/mispackaged skill rather than clearly malicious code, but treat it with caution and request a complete, verifiable release before use.
Capability Analysis
Type: OpenClaw Skill
Name: openclaw-memory-system
Version: 1.0.0
The OpenClaw Memory System skill bundle is a comprehensive framework for managing agent memory, including multimodal data and project isolation. The provided files, including the installation script (scripts/install.js) and documentation (README.md, SKILL.md), show no signs of malicious intent or data exfiltration. The system includes proactive security features such as path validation to prevent directory traversal and sensitive keyword filtering for data privacy. All instructions and code logic are consistent with the stated purpose of enhancing agent memory capabilities.
Capability Assessment
Purpose & Capability
The skill's name/description promise a full multimodal memory system (many JS modules, search, versioning, Feishu integration). However the provided file list in this submission contains mostly documentation and a single scripts/install.js; the core runtime files referenced throughout SKILL.md (image-memory.js, tool-memory.js, memory-correct.js, etc.) are not actually present. This mismatch suggests the package is incomplete or mispackaged — it does not currently deliver the capabilities it advertises.
Instruction Scope
SKILL.md instructs copying many skill and config files into the OpenClaw workspace and running node tests (e.g., node skills/multimodal-memory/test-multimodal.js). Those commands are reasonable when the code exists, but because the bundle here lacks the referenced JS modules and tests, the instructions are misleading. SKILL.md also mentions configuring a Feishu webhook for notifications (optional), which would involve sending data to an external endpoint if enabled — acceptable if deliberate and controlled, but worth noting.
Install Mechanism
There is no network download/install step — the only script (scripts/install.js) creates directories under the OPENCLAW_WORKSPACE and attempts to copy config files from a path inside that workspace. This is low risk in isolation. However package.json references postinstall and a main module that are missing; the install script expects files under WORKSPACE/skills/openclaw-memory-system/, which may never exist for this bundle. The install behavior is local-only (no external URLs) but the packaging inconsistencies are concerning.
Credentials
The skill declares no required environment variables, no credentials, and no privileged config paths. SKILL.md mentions optional Feishu webhook configuration for notifications; that is optional and would require the user to supply a webhook URL. Overall requested environment/credentials are proportionate to the described functionality.
Persistence & Privilege
Flags are default (always: false, agent invocation allowed), and the install script only creates directories and copies config files inside the workspace. There's no evidence the skill requests permanent global privileges or modifies other skills' configs. This aspect appears proportionate.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install openclaw-memory-system - After installation, invoke the skill by name or use
/openclaw-memory-system - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
openclaw-memory-system v1.0.0
- Introduces a multimodal memory system supporting image/tool memory, project/agent/user isolation, and natural language correction.
- Provides granular memory isolation, including namespaces for projects, agents, and users with permission controls.
- Enables feedback-driven memory correction with version control and Feishu card support.
- Includes installation instructions, configuration templates, and testing guidelines.
- Implements security features: path validation, permission control, atomic writes, versioning, and sensitive information filtering.
Metadata
Frequently Asked Questions
What is OpenClaw 记忆系统?
支持多模态图片和工具调用记忆，细粒度项目和用户隔离，及基于自然语言的记忆修正与版本控制。 It is an AI Agent Skill for Claude Code / OpenClaw, with 202 downloads so far.
How do I install OpenClaw 记忆系统?
Run "/install openclaw-memory-system" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is OpenClaw 记忆系统 free?
Yes, OpenClaw 记忆系统 is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does OpenClaw 记忆系统 support?
OpenClaw 记忆系统 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created OpenClaw 记忆系统?
It is built and maintained by Nick-Liu1989 (@nick-liu1989); the current version is v1.0.0.
More Skills