openclaw memory opportunity

OpenClaw Memory-OS - Digital immortality service with conversation memory extraction | 数字永生服务与对话记忆自动提取

Security Scan

The skill's stated purpose (local, file-based conversation memory) is generally consistent with its instructions, but there are notable inconsistencies (claims 'no network' while recommending npm/git installation) and behavior (auto-triggered capture of conversations and recursive file collection) that warrant caution before installing or enabling autonomously.

Purpose & Capability

The name/description (local conversation-memory, digital-immortality) aligns with the actions described: local JSON storage, file collectors, and a CLI. Requiring node/npm and an npm package is reasonable for an npm-based CLI. However, the SKILL.md asserts 'no network activity' at runtime but the installation instructions recommend 'npm install -g' or 'git clone' which require network access — a discrepancy between claimed runtime behavior and installation steps.

Instruction Scope

Runtime instructions include an AUTO-TRIGGER that will automatically extract and store information when conversational trigger words are used; this implies the agent may capture and persist user utterances without an explicit per-item confirmation unless the user disables auto-trigger. The CLI also supports recursive directory scanning and bulk collect operations which can read large numbers of user files (including sensitive ones) if broad paths are supplied. The SKILL.md repeatedly advises careful use, but the instructions grant broad discretion to read and store arbitrary files and conversation content.

Install Mechanism

No platform install spec in the registry metadata, but SKILL.md requires installing an npm package (openclaw-memory-opportunity) and/or git-cloning the repo. Fetching via npm or git involves network downloads; the package is referenced with a GitHub repo and a commit hash (a positive signal), but installing arbitrary npm packages is moderate risk and the instructions rely on network-accessed third-party code despite claiming the runtime is offline.

Credentials

The skill declares no required environment variables, no API keys, and no required config paths. This is proportionate to the stated current capabilities (local-only storage and collection). There is no request for unrelated credentials or secrets in the manifest or SKILL.md.

Persistence & Privilege

The skill does not require 'always: true' and is user-invocable; however auto-trigger behavior plus autonomous invocation (platform default) means the skill can automatically persist memories to ~/.memory-os/ during conversations. This is coherent with its purpose but increases blast radius if you enable autonomous invocation or the auto-trigger feature without restrictions.

What to consider before installing

This skill is mostly consistent with being a local, file-based memory collector, but take these precautions before installing or enabling auto-trigger: 1) Understand installation will likely download code (npm / git) — inspect the package source (the referenced GitHub repo and commit) before installing. 2) Test in a sandbox or VM first and monitor network activity during install/use. 3) Disable AUTO-TRIGGER until you are comfortable; verify exactly what the skill will persist when it detects 'remember' / '记住' triggers. 4) When using 'collect', pass narrow explicit paths and avoid broad directories (e.g., do not point it at your entire home directory). 5) Inspect the stored JSON files under ~/.memory-os/memories/ and remove them (rm -rf ~/.memory-os/) if you no longer want the data. 6) If you need stronger protections, consider encrypting the store or only installing from a vetted local copy of the repo rather than npm. If you want me to, I can: (a) list the exact files to review in the GitHub repo, (b) draft a safe install checklist, or (c) help build a minimal test command set to validate the 'no-network' claim.

OpenClaw Memory-OS

English | 中文

⚠️ Security & Privacy Notice (v0.1.2)

Current Version Status:

• ✅ 100% Local Storage - All data stored in ~/.memory-os/
• ✅ No External API Calls - Zero network activity
• ✅ No API Keys Required - Works completely offline
• ✅ Manual Collection Only - No automatic background scanning
• ✅ Conversation Memory - Auto-extract from natural language (NEW!)
• ⚠️ Future Features Planned - Semantic search and LLM features NOT yet implemented

What v0.1.2 Does:

• ✅ Local file-based memory storage (JSON)
• ✅ Basic keyword search (local)
• ✅ Batch file collection CLI (batch import from directories)
• ✅ NEW: Conversation memory extraction (auto-extract names, dates, events)
• ✅ NEW: Auto-trigger support (works with "记住..." or "remember...")
• ✅ Recursive directory scanning
• ✅ Automatic file type detection (TEXT vs CODE)
• ✅ Timeline and stats (local computation)

What v0.1.2 Does NOT Do:

• ❌ No AI embeddings
• ❌ No LLM calls
• ❌ No external API usage
• ❌ No automatic background collection
• ❌ No semantic search (planned for v0.2.0+)

Data Control:

• Your data: ~/.memory-os/
• You control: What to collect, when to collect
• You own: All data files (JSON format, human-readable)
• You delete: rm -rf ~/.memory-os/ removes everything

Recommended Safe Usage:

1. Test in sandboxed environment first
2. Review what files will be collected before running collect commands
3. Use explicit paths - avoid broad patterns like ~/Documents
4. Inspect collected data in ~/.memory-os/data/memories/
5. Disable AUTO-TRIGGER in production until you're comfortable

---

Digital immortality service and cognitive continuity infrastructure.

AI-powered personal memory management system for capturing, storing, and intelligently retrieving your digital memories.

Installation

Step 1: Install via ClawHub (Recommended)

# Install the skill
clawhub install openclaw-memory-opportunity

Step 2: Install the npm package

# Global installation
npm install -g openclaw-memory-opportunity

# Or from source
git clone https://github.com/ZhenRobotics/openclaw-memory-opportunity.git
cd openclaw-memory-opportunity
npm install
npm run build
npm link

Step 3: Initialize Memory-OS

# Initialize (creates ~/.memory-os/)
openclaw-memory-opportunity init

# Configure (optional)
openclaw-memory-opportunity config set owner.name "Your Name"
openclaw-memory-opportunity config set owner.email "[e-mail]"

Step 4: Collect Your First Memories

# Create test directory with sample files
mkdir -p ~/test-memories
echo "My first note" > ~/test-memories/note1.txt
echo "# Learning Log" > ~/test-memories/log.md

# Collect memories from directory
openclaw-memory-opportunity collect --source ~/test-memories/

# Verify collection
openclaw-memory-opportunity status
openclaw-memory-opportunity search "first"

Security Check:

# Verify data location
ls -la ~/.memory-os/memories/

# Inspect collected memories
cat ~/.memory-os/memories/*.json | head -20

---

Usage

When to Use This Skill

AUTO-TRIGGER (New in v0.1.2 - Conversation Memory):

The skill now automatically responds to memory-related conversations:

Trigger Keywords (Chinese):

• 记住 - "记住我的名字：刘小容"
• 保存 - "帮我保存这个信息"
• 记录 - "记录今天的会议内容"

Trigger Keywords (English):

• remember - "Remember my name is Liu Xiaorong"
• save to memory - "Save this to memory"
• keep in mind - "Keep in mind that..."

How It Works:

1. You mention "记住..." or "remember..." in conversation
2. The skill automatically extracts key information
3. Stores it in your local memory database
4. Confirms what was remembered

Example Usage:

User: 记住我的名字：刘小容
Agent: ✅ 已记住
       姓名: 刘小容
       置信度: 80%

User: Remember that the project deadline is 2026-04-01
Agent: ✅ Remembered
       Date: 2026-04-01
       Event: project deadline
       Confidence: 90%

MANUAL TRIGGER (For batch file operations):

Use explicit commands when you want to:

• Batch import files: "Collect memories from ~/my-notes/"
• Search your memory database: "Search my memories for 'project planning'"
• View statistics: "Show memory status"

DO NOT USE when:

• Simple reminders or todos (use task management)
• Real-time collaboration (use chat tools)
• Sensitive data without review (all data is local, but be mindful)

---

Core Features

v0.1.1 (Current):

• ✅ Local Storage - JSON-based, in ~/.memory-os/
• ✅ Batch File Collection - Import entire directories with progress display
• ✅ Automatic Type Detection - Distinguishes CODE from TEXT files
• ✅ Recursive Scanning - Processes subdirectories automatically
• ✅ Basic Search - Keyword and tag-based (local)
• ✅ Timeline - Temporal tracking of memories
• ✅ Privacy-First - No cloud, no external APIs
• ✅ Extensible - Modular architecture for future features

Planned for Future Versions:

• ⏳ Semantic Search - AI-powered (requires API key)
• ⏳ Knowledge Graph - Automatic relations (requires API key)
• ⏳ Cognitive Chat - LLM integration (requires API key)

---

Security Best Practices

1. Test in Isolated Environment

# Create test user or use VM
# Install in test environment first
npm install -g openclaw-memory-opportunity

# Initialize
openclaw-memory-opportunity init

# Create test data
mkdir ~/test-memories
echo "Test note 1" > ~/test-memories/note1.txt
echo "Test note 2" > ~/test-memories/note2.md

# Collect from test directory
openclaw-memory-opportunity collect --source ~/test-memories/

2. Review Collected Data

# Check what was collected
ls ~/.memory-os/memories/

# Read individual memory files
cat ~/.memory-os/memories/*.json | jq '.'

# View statistics
openclaw-memory-opportunity status

3. Control Collection Scope

# ✅ Good: Specific directory
openclaw-memory-opportunity collect --source ~/my-project-notes/

# ✅ Good: With exclusions
openclaw-memory-opportunity collect --source ~/Documents/ --exclude node_modules .git dist

# ⚠️ Caution: Broad scope
openclaw-memory-opportunity collect --source ~/Documents/

# ❌ Dangerous: System-wide
openclaw-memory-opportunity collect --source ~/  # DON'T DO THIS

4. Data Retention & Deletion

# View all memories
openclaw-memory-opportunity status

# Search for specific content
openclaw-memory-opportunity search "keyword"

# Delete specific memory
rm ~/.memory-os/memories/<memory-id>.json

# Complete removal
rm -rf ~/.memory-os/

5. Network Traffic Verification

# v0.1.1 should have ZERO network traffic
# Monitor with:
sudo tcpdump -i any port 443 or port 80 &
openclaw-memory-opportunity collect --source ~/test-data/
# Should see NO external connections

---

Conversation Memory Usage (v0.1.2+)

Quick Start with Conversation Memory

Command Line:

# Chinese example
openclaw-memory-opportunity remember "记住我的名字：刘小容"

# English example
openclaw-memory-opportunity remember "Remember my name is Liu Xiaorong"

# Complex information
openclaw-memory-opportunity remember "记住：项目截止日期是2026年4月1日，负责人是张三"

In Claude Conversation:

You: 记住我的名字：刘小容

Claude: ✅ 已记住！
        姓名: 刘小容
        存储位置: ~/.memory-os/memories/
        类型: TEXT

What Gets Extracted Automatically

The system intelligently extracts:

• Names: "我的名字是刘小容" → extracts "刘小容"
• Dates: "截止日期2026-04-01" → extracts "2026-04-01"
• Events: "会议：讨论Q2规划" → extracts "讨论Q2规划"
• Facts: Any other important information

Supported Languages

• Chinese (中文): 记住、保存、记录、存储
• English: remember, save, store, keep, note, record

---

Agent Usage Guide

Important Notes

NEW in v0.1.2 - Conversation Memory:

• Auto-extracts information from conversations
• Works with Chinese and English
• Stores locally in ~/.memory-os
• No manual file operations needed

CRITICAL for v0.1.1:

• This version is local-only
• No AI embeddings or LLM features active
• All operations happen on your machine
• No credentials needed
• CLI collect command is now fully functional

Package Name: When importing, use openclaw-memory-opportunity:

import { MemoryOS, MemoryType } from 'openclaw-memory-opportunity';

CLI Name: When using CLI, use openclaw-memory-opportunity:

openclaw-memory-opportunity init
openclaw-memory-opportunity collect --source ~/specific-folder

Pattern 1: Save Memory (Local Only)

import { MemoryOS, MemoryType } from 'openclaw-memory-opportunity';

const memory = new MemoryOS({
  storePath: '~/.memory-os'  // Local storage
});
await memory.init();

// Save text memory (local JSON file)
await memory.collect({
  type: MemoryType.TEXT,
  content: 'User prefers TypeScript',
  metadata: {
    tags: ['preference'],
    source: 'manual',
  },
});

Pattern 2: Search Memory (Local Only)

// Basic keyword search (no AI)
const results = await memory.search({
  query: 'TypeScript',  // Simple text matching
  limit: 5,
});

// Tag-based search
const tagResults = await memory.search({
  tags: ['preference'],
});

Pattern 3: Timeline Query (Local Only)

// Query local timeline
const timeline = await memory.timeline({
  date: new Date('2024-03-01'),
  range: 'day',
});

---

CLI Commands

Basic Operations (All Local)

# Initialize (creates local directory)
openclaw-memory-opportunity init

# Collect from specific directory (NEW in v0.1.1 - fully functional)
openclaw-memory-opportunity collect --source ~/my-notes/

# Collect with options
openclaw-memory-opportunity collect --source ~/Documents/ --exclude node_modules .git
openclaw-memory-opportunity collect --source ~/code/ --recursive

# Search locally
openclaw-memory-opportunity search "keyword"
openclaw-memory-opportunity search --type text "programming notes"

# Status (shows total memories, type breakdown)
openclaw-memory-opportunity status

Security Commands

# Inspect data location
openclaw-memory-opportunity status

# View statistics (local computation)
openclaw-memory-opportunity stats

# Export data (local file copy)
openclaw-memory-opportunity export ~/backup/

# Complete removal
rm -rf ~/.memory-os/

---

Configuration

v0.1.0 Configuration (No API keys needed):

{
  "storage": {
    "path": "~/.memory-os/data",
    "backend": "local"
  },
  "collectors": {
    "auto": false,  // Manual only
    "sources": [],  // Must be explicitly set
    "exclude": ["node_modules", ".git", ".env"]
  },
  "privacy": {
    "encryption": false,
    "shareStats": false
  }
}

Future Configuration (v0.2.0+, when AI features are implemented):

{
  "embedding": {
    "provider": "openai",  // Will require API key
    "apiKey": "${OPENAI_API_KEY}"
  },
  "llm": {
    "provider": "anthropic",  // Will require API key
    "apiKey": "${ANTHROPIC_API_KEY}"
  }
}

---

Known Limitations (v0.1.1)

1. No AI Features - Semantic search and LLM features not implemented
2. Basic Search Only - Simple keyword/tag matching (but works well with collected files)
3. Manual Collection - No automatic background scanning
4. No Encryption - Data stored as plain JSON (can enable manually)
5. No Multi-user - Single-user local storage only
6. Limited Config Commands - Config management partially implemented

---

Roadmap & Future Security Considerations

v0.2.0 (Planned) - AI Features

Will introduce:

• Semantic search (requires OpenAI/Anthropic API key)
• Embeddings generation (data sent to external API)
• LLM-powered insights

Security measures planned:

• Explicit API key configuration
• User consent for each API call
• Local-only mode option
• Encryption support

v0.3.0 (Planned) - Advanced Features

Will introduce:

• Cloud sync (optional)
• Encrypted storage
• Multi-device support

---

Links

• ClawHub: https://clawhub.ai/skills/openclaw-memory-opportunity
• npm: https://www.npmjs.com/package/openclaw-memory-opportunity
• GitHub: https://github.com/ZhenRobotics/openclaw-memory-opportunity
• Issues: https://github.com/ZhenRobotics/openclaw-memory-opportunity/issues
• Security: https://github.com/ZhenRobotics/openclaw-memory-opportunity/blob/main/SECURITY.md

---

License

MIT-0 License

---

OpenClaw Memory-OS (中文)

English | 中文

⚠️ 安全与隐私声明（v0.1.0 MVP 版本）

当前版本状态：

• ✅ 100% 本地存储 - 所有数据存储在 ~/.memory-os/data/
• ✅ 无外部 API 调用 - 零网络活动
• ✅ 无需 API 密钥 - 完全离线工作
• ✅ 仅手动收集 - 无自动后台扫描
• ⚠️ 未来功能计划中 - 语义搜索和 LLM 功能尚未实现

v0.1.0 能做什么：

• ✅ 本地文件记忆存储（JSON 格式）
• ✅ 基本关键词搜索（本地）
• ✅ 文件采集（仅手动触发）
• ✅ 时间线和统计（本地计算）

v0.1.0 不能做什么：

• ❌ 无 AI 向量化
• ❌ 无 LLM 调用
• ❌ 无外部 API 使用
• ❌ 无自动后台收集
• ❌ 无语义搜索（计划 v0.2.0+）

数据控制：

• 您的数据：~/.memory-os/data/
• 您控制：收集什么、何时收集
• 您拥有：所有数据文件（JSON 格式，人类可读）
• 您删除：rm -rf ~/.memory-os/ 删除所有内容

推荐安全使用：

1. 先在沙盒环境测试
2. 运行收集命令前检查将收集哪些文件
3. 使用明确路径 - 避免 ~/Documents 等广泛模式
4. 检查收集的数据 在 ~/.memory-os/data/memories/
5. 禁用自动触发 在生产环境，直到您熟悉系统

---

数字永生服务 | 认知延续基础设施

AI 驱动的个人记忆管理系统，用于捕获、存储和智能检索您的数字记忆。

安装

[安装步骤与英文版相同]

使用场景

手动触发（v0.1.0 推荐）：

显式使用时：

• 保存特定信息："保存到记忆：..."
• 检索特定信息："搜索我的记忆..."
• 从特定文件收集："从 ~/my-notes/ 收集记忆"

自动触发（⚠️ 谨慎使用）：

关键词：memory、remember、recall、记忆、回忆、记住、保存

⚠️ 安全建议：

• 在生产环境禁用自动触发
• 手动批准每个收集操作
• 定期检查收集的数据

---

核心功能

v0.1.0 MVP（当前）：

• ✅ 本地存储 - JSON 格式，在 ~/.memory-os/data/
• ✅ 手动收集 - 从特定文件/目录
• ✅ 基本搜索 - 关键词和标签（本地）
• ✅ 时间线 - 记忆的时间追踪
• ✅ 隐私优先 - 无云端，无外部 API
• ✅ 可扩展 - 模块化架构用于未来功能

未来版本计划：

• ⏳ 语义搜索 - AI 驱动（需要 API 密钥）
• ⏳ 知识图谱 - 自动关系（需要 API 密钥）
• ⏳ 认知对话 - LLM 集成（需要 API 密钥）

---

安全最佳实践

[安全最佳实践与英文版相同]

---

链接

• ClawHub: https://clawhub.ai/skills/openclaw-memory-opportunity
• npm: https://www.npmjs.com/package/openclaw-memory-opportunity
• GitHub: https://github.com/ZhenRobotics/openclaw-memory-opportunity
• 问题反馈: https://github.com/ZhenRobotics/openclaw-memory-opportunity/issues
• 安全: https://github.com/ZhenRobotics/openclaw-memory-opportunity/blob/main/SECURITY.md

---

许可证

MIT-0 License

---

Memory-OS v0.1.2 - 100% Local, 0% Cloud, Your Data, Your Control

Version: 0.1.2 | Verified Commit: 28a1a92 | Status: Production-Ready with Conversation Memory