MemTensorOfficial·6 Agent Skills Included

MemOS

Give AI agents persistent long-term memory that grows

Adds persistent long-term memory to AI agents and LLM applications through a unified store, search, and manage API. Eliminates forgotten conversations, repeated questions, and black-box embedding stores with inspectable graph-based memory cubes. Supports text, images, tool traces, and preferences with async scheduling, feedback correction, and multi-agent sharing. Runs as a Python library, self-hosted FastAPI service, or local plugin for OpenClaw, Hermes, and DeepSeek Harness.

10,655|987

↗

npx skills add MemTensor/MemOS --all -g -y

Available:

Gives the AI agent the project architecture, provider registry, testing rules, and behavior boundaries it must follow when working on or extending the MemOS memory system.

All Skills in This Repository (6)

Pure Emerald Level Indicators

📦 In Repo

MemTensor

memos-memory-guide

Search and write MemOS Local memories to recall past conversations and discover skills.

browserwing-executor

Automate browser interactions via HTTP API for navigation, element manipulation, and data extraction.

browserwing-admin

Manage BrowserWing browser automation via RESTful API operations.

ask-user-question

Display graphical questions with single or multi-select options and custom text input.

dev-browser

Automate browser interactions and workflows using Playwright.

safe-file-deletion

Require explicit user confirmation before executing file deletion commands.

Official

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install MemOS?▼

Run `npx skills add MemTensor/MemOS --all -g -y` in your terminal to install all skills in this suite globally.

What does MemOS do for AI agents?▼

MemOS gives agents persistent long-term memory, so they can recall past conversations, user preferences, and task history across sessions instead of starting fresh every time.

Can I self-host MemOS on my own servers?▼

Yes. You can run it with Docker Compose, which starts the MemOS API together with Neo4j and Qdrant, keeping all data on your own infrastructure.

Does MemOS work with OpenClaw and other agent frameworks?▼

Yes. MemOS provides official local and cloud plugins for OpenClaw, Hermes, and DeepSeek Harness, plus a standard API any agent framework can call.

Can MemOS run fully offline without cloud accounts?▼

Yes. The local plugin stores everything in an on-device SQLite database with a local embedding model, requiring no API keys or registration.