← 返回 Skills 市场
458
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install xiucheng-memory-manager
功能描述
Long-term memory management system for OpenClaw agents. Automatically archives conversations, organizes memories, and provides semantic search capabilities.
使用说明 (SKILL.md)
Memory Manager
A robust long-term memory management system for OpenClaw agents that automatically archives conversations and enables intelligent memory retrieval.
Features
- 📝 Automatic Conversation Logging: Saves dialogues to daily memory files
- 🔍 Semantic Search: Retrieve relevant memories using natural language queries
- 📅 Time-based Organization: Memories organized by date (YYYY-MM-DD.md)
- 🗂️ Long-term Storage: Curated memories in MEMORY.md
- 🔗 Integration Ready: Works with existing agent workflows
Installation
clawhub install memory-manager
Usage
Automatic Mode
The skill automatically saves conversations during agent operation.
Manual Memory Search
from memory_manager import MemoryManager
mm = MemoryManager()
results = mm.search_memory("project discussions about KPI")
Daily Log Structure
memory/
├── 2026-03-07.md
├── 2026-03-06.md
└── ...
Configuration
| Setting | Default | Description |
|---|---|---|
| memory_dir | ./memory/ | Daily memory files location |
| memory_file | ./MEMORY.md | Long-term memory file |
| auto_save | true | Auto-save conversations |
Requirements
- Python 3.8+
- OpenClaw workspace structure
License
MIT License - Contribute back to the community!
安全使用建议
This package is a small, local file-based memory helper, but its docs overstate capabilities. Before installing or enabling it: 1) Note that 'semantic search' is not implemented — the search is simple keyword matching. 2) The README claims automatic saving, but the code provides a MemoryManager class you must call; there is no automatic agent hook. 3) The code defaults to /root/.openclaw/workspace (not the ./memory/ path shown in docs) and will create/append files there — update the workspace path or run in an isolated workspace if you don't want persistent files in that location. 4) Review the Python file yourself; it contains no network calls or secret access, but any skill that writes local files can store sensitive conversation contents. If you need true semantic search, ask the author for embedding/vector search integration or add it yourself; if you need automatic capture, require explicit agent integration hooks. Install in a test environment first and confirm behavior.
功能分析
Type: OpenClaw Skill
Name: xiucheng-memory-manager
Version: 1.0.0
The memory-manager skill is a straightforward utility for logging and searching conversation history locally within the OpenClaw workspace. The code in memory_manager.py performs standard file I/O operations to manage markdown files and lacks any indicators of data exfiltration, remote execution, or malicious intent.
能力评估
Purpose & Capability
The name/description promise automatic conversation archiving and semantic search. The included code implements local file-based logging and simple keyword matching, not semantic (embedding/vector) search. The SKILL.md implies 'automatic' saving during agent operation, but the code exposes an API/class (MemoryManager) and CLI and does not provide an integration hook to automatically capture agent conversations. This is a meaningful feature mismatch.
Instruction Scope
SKILL.md claims automatic saving (auto_save: true) and integration-ready behavior but gives no concrete integration instructions or agent hook. The code will create and write files in a workspace path and read all *.md files in that directory for searches; those file I/O operations are within scope for a memory manager but SKILL.md is vague about when/how data is collected, which grants broad discretion to any agent code that uses the library.
Install Mechanism
No install spec; instruction-only plus a small Python file. Nothing is downloaded at install time and no executables are injected by an installer. Risk from install mechanism is low.
Credentials
The skill requests no credentials or env vars. However there is a documentation/code inconsistency: SKILL.md lists defaults like './memory/' and './MEMORY.md' while the MemoryManager default workspace in code is '/root/.openclaw/workspace', which causes the skill to read/write under that path by default. This is not a credential request but it changes where files are stored and may surprise users.
Persistence & Privilege
The skill is not always-enabled and does not request elevated privileges. It writes to its own workspace directory and does not modify other skills or system-wide settings. Autonomous invocation is allowed by platform default but not a separate concern here.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install xiucheng-memory-manager - 安装完成后，直接呼叫该 Skill 的名称或使用
/xiucheng-memory-manager触发 - 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-manager for OpenClaw agents.
- Automatically archives conversations and organizes them by date.
- Enables semantic search to retrieve relevant memories using natural language.
- Curates important memories in a dedicated long-term storage file.
- Seamlessly integrates with existing agent workflows.
- Simple configuration with support for auto-save and custom storage locations.
元数据
常见问题
Memory Manager 是什么？
Long-term memory management system for OpenClaw agents. Automatically archives conversations, organizes memories, and provides semantic search capabilities. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 458 次。
如何安装 Memory Manager？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install xiucheng-memory-manager」即可一键安装，无需额外配置。
Memory Manager 是免费的吗？
是的，Memory Manager 完全免费（开源免费），可自由下载、安装和使用。
Memory Manager 支持哪些平台？
Memory Manager 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 Manager？
由 xiucheng（@xiucheng）开发并维护，当前版本 v1.0.0。
推荐 Skills