Token导航 LogoToken导航TokenDH.com
开发需要联网clawhub未标认证来源可访问clear审计提醒

memory-auto自动记忆

Agent Skill

memory-auto 用于补充开发相关能力,适合在 OpenClaw 中需要让 Agent 承接开发相关任务时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

8,936

周安装

358

GitHub Stars

公开资料未说明

下载量

2,893
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:memory-auto(自动记忆)
来源仓库:https://github.com/yanshijun607-png/memory-auto
安装命令:
openclaw skills install memory-auto
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install memory-auto

简介

自动将带有关键字亮点和可选 AI 摘要的每日聊天日志归档到有组织的内存文件中,无需手动设置。

SKILL.md

openclaw-memory-auto

Automatic memory archiving for OpenClaw. Works for everyone.

Features

  • Zero Configuration: Works out of the box with smart defaults
  • Cross-Platform: Pure TypeScript, runs on Windows/Mac/Linux
  • Universal: Language-agnostic keywords, customizable to your workflow
  • Non-Intrusive: Runs on startup or cron, no manual work needed
  • Respectful: Only archives what matters, not every chat

Installation

npm install openclaw-memory-auto

Quick Start

Add to your openclaw.config.js:

import memoryAuto from 'openclaw-memory-auto';

export default {
  plugins: [memoryAuto]
};

That's it! The plugin will:

  • ✅ On every agent startup, archive yesterday's chat
  • ✅ Highlight key moments using smart keyword matching
  • ✅ Store daily logs in memory/YYYY-MM-DD.md
  • ✅ Prepare prompts for AI-assisted MEMORY.md updates

Configuration (Optional)

Customize behavior in openclaw.config.js:

import memoryAuto from 'openclaw-memory-auto';

export default {
  plugins: [memoryAuto],
  
  memoryAuto: {
    // Add your own keywords (case-insensitive)
    keywords: [
      '任务', '部署', 'API', 'token', 'OpenClaw',
      'task', 'deploy', 'config', 'password', 'secret'
    ],
    
    // Custom paths (relative to workspace)
    paths: {
      memoryDir: 'memory',
      logsDir: 'logs'
    },
    
    // Schedule settings
    schedule: {
      archiveHour: 9,      // 9 AM local time
      checkOnStartup: true // always check on startup
    },
    
    // Enable AI refinement (beta)
    refine: {
      enabled: true,
      model: 'stepfun/step-3.5-flash:free',
      prompt: `Extract long-term memories from today's log.`
    }
  }
};

How It Works

  1. On Startup: Plugin checks if yesterday's daily log exists
  2. If Missing: Scans all transcript files (agents/*/sessions/*.jsonl)
  3. Time Filter: Extracts only yesterday's messages (local timezone aware)
  4. Highlighting: Matches messages against your keyword list
  5. Archive: Writes memory/YYYY-MM-DD.md with summary and highlights
  6. Refine (optional): Calls AI to update MEMORY.md with structured memory

File Structure

After running:

workspace/
├── memory/
│   ├── 2026-03-11.md      # Auto-generated daily log
│   └── .2026-03-11.archived  # Marker (prevents re-archive)
├── logs/
│   ├── memory-archive.log
│   └── last_refine_prompt.txt  # For AI refinement
└── MEMORY.md              # Your long-term memory (manually or auto-updated)

Custom Keywords

Add to your config:

memoryAuto: {
  keywords: [
    // Defaults already included - just add yours
    '项目名', '客户', '截止日期',
    'project', 'client', 'deadline'
  ]
}

Keywords are case-insensitive and support Unicode (中文, emoji, etc.).

AI Refinement

Enable in config to automatically update MEMORY.md:

memoryAuto: {
  refine: {
    enabled: true,
    model: 'stepfun/step-3.5-turbo:free',
    prompt: `Extract: skills learned, project milestones, user preferences, important data, inside jokes.`
  }
}

The AI will append structured sections to MEMORY.md daily.

FAQ

Q: Does it archive everything? A: No. Only messages containing your keywords (default: 30+ common terms). This keeps logs meaningful and privacy-friendly.

Q: Where are transcripts stored? A: OpenClaw automatically saves sessions to agents/*/sessions/*.jsonl. The plugin reads these.

Q: Can I run on a schedule instead of startup? A: Yes. Use OpenClaw's heartbeat cron:

agents: {
  defaults: {
    heartbeat: {
      tasks: [{
        name: 'memory-auto-archive',
        cron: '0 9 * * *', // daily 9 AM
        command: 'npx openclaw-memory-auto --archive'
      }]
    }
  }
}

Q: I have multiple agents. Will it work? A: Yes. Scans all agents/*/sessions/ directories.

Development

git clone https://github.com/openclaw-plugins/memory-auto.git
cd memory-auto
npm install
npm run build

License

MIT - Free for personal and commercial use.

Contributing

PRs welcome! Please read CONTRIBUTING.md first.


Made with 🦞 by the OpenClaw Community

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

75.29%
按下载量换算2,178

安全审计

VirusTotal

可疑

ClawScan

可疑

Static analysis

可疑

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills