ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified productivity Safety 4/5

memory-sync-enhanced

增强版记忆系统 - Ebbinghaus 遗忘曲线 + Hebbian 共现图

Why use this skill?

Upgrade your OpenClaw agent with an advanced memory system using Ebbinghaus forgetting curves and Hebbian co-occurrence graphs for smarter, context-aware recall.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/guohongbin-git/memory-sync-enhanced
Or

What This Skill Does

The memory-sync-enhanced skill provides a sophisticated, dual-layered memory architecture for the OpenClaw AI agent. By synthesizing the Ebbinghaus Forgetting Curve with a Hebbian Co-occurrence graph, the system mimics biological memory retention. The first layer, a CortexGraph, handles semantic searching enriched with decay factors based on usage frequency and the time elapsed since the memory was created. The second layer uses a persistent SQLite-based graph database to track relationships between distinct memory nodes, allowing the AI to bridge disparate topics—such as music preferences and programming syntax—if they are frequently retrieved together. This mechanism ensures that the agent "forgets" trivial information over time while reinforcing critical insights and frequent associations.

Installation

To integrate this memory system, ensure you have the OpenClaw CLI installed and initialized in your workspace. Execute the following command in your terminal:

clawhub install openclaw/skills/skills/guohongbin-git/memory-sync-enhanced

Once installed, verify the setup by running ./scripts/stats.sh to observe your current memory distribution and baseline decay status.

Use Cases

This skill is ideal for power users and developers who require a context-aware AI that evolves alongside their workflow. Use it to:

  1. Maintain a long-term knowledge base from daily coding logs and research notes.
  2. Automatically surface related project files based on recent task associations.
  3. Automate the cleanup of obsolete information via the integrated garbage collection script.
  4. Build complex neural bridges between professional projects and personal knowledge management.

Example Prompts

  1. "OpenClaw, find all memory nodes related to 'Quant Trading' and boost them based on my recent interactions in the coding notebook."
  2. "Perform a garbage collection on my current memory store; delete any entries with a decay score below 0.1 to clear clutter."
  3. "Show me the current co-occurrence links between my 'Music' tags and my 'Project-Alpha' development logs."

Tips & Limitations

  • Monitor Decay: Regularly use ./scripts/stats.sh to ensure your "Healthy" memory count remains stable. If too many items fall into the "Danger zone," consider increasing the strength parameter on important memories.
  • Promotion: Use the ./scripts/promote.sh command to manually elevate critical STM entries to LTM status, ensuring they remain protected from the aggressive 3-day half-life decay.
  • Performance: While SQLite is highly efficient for the Co-occurrence graph, ensure regular backups of your memory/ directory to prevent data loss during system upgrades.

Metadata

Stars2387
Views0
Updated2026-03-09
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-guohongbin-git-memory-sync-enhanced": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#memory#knowledge-graph#productivity#automation#personal-ai
Safety Score: 4/5

Flags: file-write, file-read, data-collection