Micro Memory

智能记忆系统，支持记忆添加、搜索、标签管理、强度追踪、复习提醒及统计健康报告。

Security Scan

Micro Memory is internally consistent: its description (local intelligent memory/CLI) matches the code and runtime instructions, it only reads/writes files in its own store directory and does not request credentials or contact external services.

Purpose & Capability

Name/description (intelligent local memory, search, spaced repetition, health reports) align with the included TypeScript/JS implementation. The code implements CLI commands (add/search/list/review/health/etc.), local JSON/markdown storage, linking, compression and archiving — all expected for this purpose. No unrelated credentials, binaries, or cloud integrations are requested.

Instruction Scope

SKILL.md describes local CLI usage and the repo structure; the runtime code follows those instructions and only executes local CLI operations. The instructions and code reference only the skill's store files and current working directory for exports. There are minor documentation inconsistencies (SKILL.md lists version 4.0.1 while registry is 4.0.2), but nothing expands scope to access system-wide secrets or remote endpoints.

Install Mechanism

No install spec is declared (instruction-only at registry level), and the package includes source and compiled dist files. SKILL.md suggests using npm install and npm run build if you want to build locally; there is no external download or installer that pulls code from an arbitrary URL. Risk from installation is standard for running a Node CLI (you must run npm locally to build/run if you choose).

Credentials

The skill requires no environment variables or external credentials. It reads/writes only to its own store directory (exports may be written to the current working directory). No env access beyond that is present in code. Users should avoid storing sensitive secrets in memories because those are persisted to disk and exported as JSON/CSV.

Persistence & Privilege

always is false and the skill does not request elevated system privileges. It persists data under the skill's store/ path and writes store.md and JSON files; this is expected behavior. The skill can be invoked autonomously by the agent (default), which is normal — there is no cross-skill configuration or system-level modification.

Assessment

This skill appears to do what it says: a local, file-backed memory CLI. Before installing, note: (1) it stores all memories in files under the skill's store/ directory and will write exported JSON/CSV to your current working directory — do not put passwords or private keys into memories you want kept secret; (2) SKILL.md recommends running npm install / npm run build if you want to build locally — standard Node workflow; (3) the skill does not contact external services or require credentials, so network exfiltration is not present in the reviewed code, but you should still inspect any future updates for new network calls or env var usage; (4) the documentation version number differs slightly from the registry metadata (minor); and (5) because agent autonomous invocation is allowed by default, be aware that trigger phrases could cause the skill to add/search/list memories without explicit manual CLI execution — if that is undesirable, disable autonomous invocation in your agent settings.

Micro Memory V4 — Your Second Brain

🧠 Not just a note-taking tool, but your external brain that truly remembers.

不只是笔记工具，而是真正会"记忆"的外部大脑。

What is Micro Memory? | 什么是 Micro Memory？

Micro Memory is an intelligent memory system inspired by how the human brain works:

• 🧠 Like your brain, memories fade over time without reinforcement
• 🔗 Like your brain, related concepts form neural networks
• 📅 Like your brain, spaced repetition strengthens retention
• 🏥 Unlike your brain, it gives you health reports and analytics

Micro Memory 是一个智能记忆系统，灵感来自人脑的工作方式：

• 🧠 像大脑一样，记忆会随时间衰减，需要强化巩固
• 🔗 像大脑一样，相关概念形成神经网络关联
• 📅 像大脑一样，间隔重复加强记忆留存
• 🏥 ** unlike大脑**，它能给你健康报告和分析

---

Why Micro Memory? | 为什么选择 Micro Memory？

Feature	Traditional Notes	Micro Memory
Storage	Static files	Living memories with strength
Search	Simple text	Multi-keyword + Regex + Fuzzy
Organization	Folders/tags	Semantic networks
Retention	Manual review	Spaced repetition scheduling
Health	No insight	Memory health reports
Compression	None	TF-IDF smart compression

特性	传统笔记	Micro Memory
存储	静态文件	有强度的活记忆
搜索	简单文本	多关键词 + 正则 + 模糊
组织	文件夹/标签	语义网络
留存	手动复习	间隔重复调度
健康	无洞察	记忆健康报告
压缩	无	TF-IDF 智能压缩

---

Core Features | 核心功能

📝 Smart Memory Creation | 智能记忆创建

memory add "Learned about TypeScript decorators"
memory add "Important meeting with client" --tag=work --type=longterm --importance=5
memory add "Related concept" --tag=study --link=1,2

🔍 Advanced Search | 高级搜索

memory search "TypeScript"                    # Multi-keyword
memory search "Type.*script" --regex          # Regex pattern
memory search "Typscript" --fuzzy             # Fuzzy matching
memory search "meeting client" --tag=work     # Filtered search

🔗 Memory Networks | 记忆网络

memory link --from=1 --to=2,3    # Connect memories
memory graph --id=1              # Visualize connections
memory graph                     # Full network view

📊 Strength System | 强度系统

Memories have 5 strength levels that decay over time:

Level	Score	Icon	Description
💎 Permanent	80-100	💎	Permanent memory
💪 Strong	60-79	💪	Strong memory
📊 Stable	40-59	📊	Stable memory
⚠️ Weak	20-39	⚠️	Weak memory - needs review
🔴 Critical	0-19	🔴	Critical - about to be forgotten

🔄 Spaced Repetition | 间隔重复

Based on SM-2 algorithm:

• Level 0 → 1 day
• Level 1 → 3 days
• Level 2 → 7 days
• Level 3 → 14 days
• Level 4 → 30 days
• Level 5+ → 60-90 days

🏥 Health Reports | 健康报告

memory health    # Overall system health
memory stats     # Detailed statistics
memory strength  # Strength distribution

📦 Compression & Archiving | 压缩与归档

memory compress --auto           # Auto-compress weak memories
memory compress --id=1           # Compress specific memory
memory archive --older_than=30   # Archive old memories
memory archive --restore=1       # Restore from archive

---

Quick Start | 快速开始

Installation | 安装

npm install
npm run build

Basic Usage | 基本用法

# Add a memory
memory add "Learned about neural networks"

# List all memories
memory list

# Search memories
memory search "neural"

# Check health
memory health

# Review due memories
memory review --today

---

Clawdbot Integration | Clawdbot 集成

As a native skill, Micro Memory auto-triggers on:

Trigger	Action
"记住..." / "记录..."	Auto-add memory
"搜索记忆..." / "找找之前..."	Auto-search
"列出记忆" / "我的记忆"	Auto-list
"记忆健康"	Show health report
"复习记忆"	Show today's reviews

---

Data Storage | 数据存储

store/
├── index.json      # Main memory index
├── links.json      # Association networks
├── reviews.json    # Review schedules
├── store.md        # Markdown backup
└── archive/        # Archived memories

---

Version | 版本

• Current | 当前: 4.0.1
• Implementation | 实现: TypeScript (Native Clawdbot Skill)

---

Architecture | 架构

micro-memory/
├── src/
│   ├── index.ts      # CLI entry
│   ├── memory.ts     # Core memory management
│   ├── strength.ts   # Strength decay system
│   ├── links.ts      # Network associations
│   ├── review.ts     # Spaced repetition
│   ├── health.ts     # Health analytics
│   ├── archive.ts    # Compression & archiving
│   └── utils.ts      # Utilities
└── dist/             # Compiled JavaScript

---

Your thoughts deserve to be remembered. Not just stored.

你的想法值得被记住，而不只是被存储。