A persistent memory layer for AI agents that enables semantic search and long-term context retention across multiple sessions.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install vanar-neutron-memory
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 vanar-neutron-memory 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 Neutron Memory?
Neutron Memory serves as a sophisticated long-term storage and retrieval system for AI agents, allowing them to maintain continuity beyond a single conversation. By integrating this into your Openclaw Skills workflow, your agent can save critical preferences, technical decisions, and user interactions into a vector-based memory bank.
Unlike traditional keyword search, Neutron Memory utilizes semantic understanding to find relevant context by meaning. This ensures that the most pertinent information is always available to the agent, enabling a personalized and highly efficient user experience that evolves as the agent learns more about your projects and workflows.
Neutron Memory Use Cases
- Persisting user coding styles and architectural preferences across different project directories.
- Building a project-specific knowledge base that an agent can query to understand previous design decisions.
- Reducing manual context setting by automatically injecting relevant past interactions into the current prompt.
- Creating a shared memory pool for multi-agent systems to synchronize knowledge and state.
How Neutron Memory Works
- Data is captured either manually via CLI commands or automatically through integrated lifecycle hooks.
- The information is sent to the Neutron API where it is processed and stored using vector embeddings.
- When the agent needs information, a semantic search query is executed to find the most relevant 'memories' based on the current conversation context.
- Relevant results are retrieved and optionally injected back into the agent's prompt to provide immediate, actionable context.
Neutron Memory Setup
To get started with this memory layer for Openclaw Skills, follow these steps:
- Generate a free API key at https://openclaw.vanarchain.com/.
- Set your environment variable:
export API_KEY=nk_your_key
- Verify the installation and connectivity using the diagnostic tool:
./scripts/neutron-memory.sh test
Neutron Memory Data Schema & Taxonomy
The skill organizes data semantically, focusing on content and intent. Below is the primary structure used for memory management:
| Property | Description | Example |
|---|---|---|
| Content | The actual text or conversation turn to be remembered | "User prefers React over Vue" |
| Title | A categorization label for the memory | "Preferences" |
| Threshold | The similarity score (0-1) for semantic matching | 0.5 |
| Limit | The maximum number of memories to recall | 10 |
Neutron Memory Advanced Features
- Auto-Recall: Automatically identifies and injects relevant memories into the agent's context before each turn using pre-tool hooks.
- Auto-Capture: Enables seamless background saving of all user and assistant turns to build a comprehensive history without manual input.
- Semantic Search Tuning: Fine-tune the relevance of recalled context using adjustable similarity thresholds and result limits.
- Diagnostic Suite: Includes a robust diagnosis script to troubleshoot API authentication, jq dependencies, and network connectivity.
SKILL.md
Loading
Related Openclaw Skills
Neutron Agent Memory
A persistent memory solution for AI agents that leverages semantic search and structured context storage to maintain state across sessions.
Android SMS Gateway
A self-hosted SMS gateway solution that turns an Android device into a private HTTP API for sending and receiving messages.
YouTube Uploader
A Python-powered automation tool for uploading videos, managing custom thumbnails, and handling OAuth2 authentication for YouTube channels.
spotify-player
A terminal-centric skill for managing Spotify playback and track searching through high-performance CLI tools.
LAN Media Server
A lightweight Node.js server that enables AI agents to share media files over a local network when standard communication channels lack inline support.
VibeTrader
VibeTrader is an AI-driven trading platform that enables users to build, backtest, and deploy automated trading bots for stocks and crypto using natural language.
METADATA
Featured*
