A high-performance durable memory system for OpenClaw agents to store, retrieve, and manage long-term project context and user preferences.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memory-memoria
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-memoria 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 (Memoria)?
Memory (Memoria) serves as the sophisticated long-term storage layer for OpenClaw agents, moving beyond simple file-based notes to a robust, external memory system. It allows agents to maintain continuity across sessions by treating Memoria as a durable memory slot, ensuring that critical project decisions, user preferences, and complex workflows are never lost.
By integrating this skill, developers can leverage advanced memory governance features like snapshots and branching. Unlike standard file memory, Memoria focuses on semantic retrieval and explicit writes, giving the user and the agent precise control over what information is deemed important enough for permanent storage within the Openclaw Skills ecosystem.
Memory (Memoria) Use Cases
- Resuming complex coding tasks after a session break by retrieving previous context.
- Managing long-term user preferences and style guidelines without re-prompting.
- Creating reversible memory branches for testing experimental workflows or risky state changes.
- Documenting durable project decisions and architectural choices for future reference.
- Correcting or purging outdated information to maintain a clean, high-precision knowledge base.
How Memory (Memoria) Works
- Initialization: At the start of a conversation, the agent uses semantic search or retrieval tools to pull relevant past context from the Memoria slot.
- Selective Storage: Rather than logging every turn, the agent identifies and stores only durable facts, significant progress, or specific user instructions.
- Active Maintenance: When errors are identified, the agent uses specialized repair tools like memory_correct or memory_forget to ensure the data remains accurate.
- State Protection: Before performing large-scale rewrites or purges, the system creates a snapshot to allow for easy rollback if necessary.
- Closure & Cleanup: Upon finishing a task, the agent stores the final outcome and purges obsolete temporary data to optimize future retrieval.
Memory (Memoria) Setup
To get started with this skill, install the plugin and initialize the Memoria environment using the following commands:
openclaw plugins install @matrixorigin/memory-memoria
openclaw plugins enable memory-memoria
openclaw memoria install
openclaw memoria verify
Memory (Memoria) Data Schema & Taxonomy
The skill organizes information into a structured taxonomy to ensure high precision in retrieval. It utilizes the following primary data structures:
| Tool Type | Purpose | Metadata Managed |
|---|---|---|
| memory_profile | Stable Preferences | User identity, coding styles, environment specs |
| memory_store | Durable Memory | Project facts, architectural decisions, task outcomes |
| Snapshots | State Recovery | Version timestamps, state hashes |
| Branches | Experimentation | Branch ID, parent ID, diff logs |
All data is managed externally to ensure it persists independently of the individual Openclaw Skills session files.
Memory (Memoria) Advanced Features
- Version Control for Memory: Support for branching and merging memory states, allowing agents to explore 'what-if' scenarios without corrupting the main knowledge base.
- Semantic Retrieval: Advanced search capabilities that go beyond keyword matching to find contextually relevant information.
- Snapshot & Rollback: Create point-in-time backups of the memory state to protect against data corruption or incorrect AI learning.
- Multi-Tool Surface: Access to a specialized suite of tools including memory_retrieve, memory_search, and memory_purge for granular data governance.
SKILL.md
Loading
Related Openclaw Skills
Refuel Record: Daxin Fleet Gas Log Automation
A specialized automation tool for Daxin Fleet to process raw refueling data into comprehensive gas summary sheets and card deduction logs.
Web Search Scraper API Skill
A professional-grade web extraction tool that transforms any website URL into structured Markdown content using the BrowserAct API.
Spc Reference Tool
A zero-latency, local reference utility that provides instant access to technical documentation, patterns, and security best practices.
Hillstone Threat Intelligence Skill
A high-performance security skill for querying IP, domain, URL, and file hash reputation data from the Hillstone threat intelligence database.
Prompt-to-Skill Router
An intelligent execution-path router that determines whether a request should be handled directly or via specialized tools.
Credit Repair Cloud
A robust integration for automating credit repair business operations, including client management, dispute processing, and affiliate tracking.
Featured*
