A sophisticated three-layer memory architecture that provides AI agents with persistent context and semantic search capabilities across sessions.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memory-mastery
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install memory-mastery using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Memory Mastery?
Memory Mastery is a robust framework designed to solve the problem of agent amnesia. By implementing a structured three-layer architecture, it ensures that your Openclaw Skills retain knowledge, preferences, and project history indefinitely. The system moves beyond simple session logs, offering a curated path from raw daily data to long-term semantic knowledge.
This skill transforms how agents interact with their environment by providing a reliable source of truth. With a focus on privacy and efficiency, it uses a combination of daily markdown logs, a curated long-term memory file, and optional vector search via the memory-core plugin to make every interaction smarter than the last.
Memory Mastery Use Cases
- Retaining project context and decisions across multiple agent restarts
- Preventing repetitive questioning by storing user preferences and past interactions
- Building a searchable knowledge base of AI-assisted development history
- Maintaining continuity in long-term research or coding tasks using Openclaw Skills
- Automating the distillation of daily activities into permanent organizational memory
How Memory Mastery Works
- The agent records raw session data and insights into append-only daily logs (L1) located in the memory directory.
- Significant insights and permanent knowledge are curated and moved into a centralized long-term memory file (L2) for persistent access.
- The system utilizes semantic vector search (L3) to find relevant context across all historical records instantly.
- Integrated maintenance scripts help distill daily logs into high-value memory entries while cleaning up outdated information.
- Custom rules within Openclaw Skills ensure the agent checks these memory layers at the start of every session.
Memory Mastery Setup
To begin using this architecture for your Openclaw Skills, follow these steps:
- Run the diagnostic audit to check your current workspace state:
bash scripts/audit.sh
- Execute the interactive setup script, providing your workspace path:
bash scripts/setup.sh /your/workspace/path
- Confirm the modifications to your AGENTS.md and HEARTBEAT.md files when prompted.
- (Optional) Install the memory-core plugin and configure an embedding API key to enable L3 vector search.
Memory Mastery Data Schema & Taxonomy
The system organizes data into a hierarchical structure within the workspace:
| File/Folder | Layer | Description |
|---|---|---|
| MEMORY.md | L2 | Long-term, curated permanent knowledge and preferences. |
| memory/ | L1 | Directory containing daily markdown logs (YYYY-MM-DD.md). |
| memory_search | L3 | Semantic index for vector-based context retrieval. |
| HEARTBEAT.md | Meta | Tracks maintenance tasks and memory health checks for Openclaw Skills. |
| AGENTS.md | Meta | Contains the core logic and rules for how agents interact with the memory. |
Memory Mastery Advanced Features
- Semantic vector search for finding memories by meaning rather than just keyword matching
- Automated maintenance helper that scans logs to suggest long-term memory updates
- Privacy-aware loading that restricts sensitive long-term memory to private sessions
- Pre-compaction triggers that prompt the agent to save critical context before it is lost
- Customizable templates for memory rules and maintenance frequency for advanced Openclaw Skills users
SKILL.md
Loading
Related Openclaw Skills
GuavaSuite
A premium 2-layer security suite for AI agents featuring $GUAVA token-gated strict protection and on-chain identity verification.
GuavaMemory
A structured episodic memory system that uses Q-value scoring to help AI agents remember successful patterns and avoid past mistakes.
GuavaGuard
A real-time runtime security monitor that identifies and warns against malicious patterns during agent execution.
whoopskill
A comprehensive CLI tool for fetching and analyzing WHOOP health metrics including sleep, recovery, HRV, and strain.
mv-pipeline
An end-to-end automated pipeline for creating professional AI-generated music videos from lyrics generation to YouTube upload.
SplitXCH Royalty Split Builder
An AI-driven utility for generating automated Chia blockchain payment split addresses from plain language descriptions.
Featured*
