A lightweight, interface-only package defining standardized abstract classes for LLM memory, search, and vector operations.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install llm-memory-interface-only
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 llm-memory-interface-only 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 LLM Memory Integration?
LLM Memory Integration is a specialized package designed to provide a consistent set of abstract base classes (ABCs) for AI memory management. Part of the broader Openclaw Skills ecosystem, this skill focuses strictly on interface definitions—Memory, Search, and Vector—ensuring that developers can build modular and swappable memory components without being tied to a specific implementation.
By separating the interface from the logic, it allows for high security and flexibility in complex AI agent architectures. This approach ensures that the core interface package remains lightweight, with zero network or file system access, making it an ideal foundation for secure enterprise AI applications.
LLM Memory Integration Use Cases
- Standardizing memory access patterns across different AI agent projects using Openclaw Skills.
- Developing private, custom implementations of vector databases while maintaining a consistent API.
- Decoupling AI search and embedding logic from core application code to improve maintainability.
- Building secure, audit-friendly AI applications where interface and implementation are strictly separated.
How LLM Memory Integration Works
- The developer identifies the required memory capabilities and imports the relevant Abstract Base Classes (MemoryInterface, SearchInterface, or VectorInterface) from the package.
- A custom implementation class is created that inherits from these interfaces, ensuring all required methods are defined.
- Implementation logic for methods like add, search, or embed is written to connect with specific backends such as vector databases or local storage.
- The AI agent interacts with these standardized methods, allowing developers to swap underlying memory providers within the Openclaw Skills framework without refactoring the agent code.
LLM Memory Integration Setup
To integrate this skill into your project, you must first define your implementation or clone the provided private implementation.
# Clone the private implementation repository
git clone https://cnb.cool/llm-memory-integrat/llm.git
Within your Python environment, you can then import the interfaces to begin your implementation:
from src.interfaces import MemoryInterface, SearchInterface, VectorInterface
LLM Memory Integration Data Schema & Taxonomy
The skill defines the following interface structures to organize data flow and metadata taxonomy:
| Interface | Core Responsibility | Methods Included |
|---|---|---|
| MemoryInterface | Memory CRUD management | add, get, update, delete, list |
| SearchInterface | Retrieval and discovery | search, hybrid_search, fts_search |
| VectorInterface | Embedding and math logic | embed, embed_single, similarity, batch_embed |
Note: As a pure interface package, no local data files are generated by the skill itself; data organization is handled by the implementation layer.
LLM Memory Integration Advanced Features
- Pure Interface Architecture: Zero implementation code ensures maximum security and zero risk of hidden subprocesses or network calls.
- Hybrid Search Support: Includes definitions for combined full-text and vector-based retrieval strategies.
- Modular Vector Operations: Standardized methods for single and batch embeddings to optimize performance in Openclaw Skills pipelines.
- Extensible Design: Designed to serve as a blueprint for multi-agent systems requiring shared or distributed memory patterns.
SKILL.md
Loading
Related Openclaw Skills
Field Research: Bilingual Academic Search Agent
A specialized AI agent skill for automated, bilingual academic literature retrieval from CNKI and Google Scholar with structured Markdown output.
Qmsg Push
A streamlined utility for sending instant QQ notifications from AI agent workflows using the Qmsg service.
Feishu All-in-One
A comprehensive integration for Feishu that centralizes messaging, cloud document collaboration, calendar scheduling, and multi-dimensional table management.
Emotion Chat Pro
An AI-powered emotional companion and high-EQ chat partner designed for role-playing, relationship advice, and workplace communication.
User Research Cookiy
A comprehensive end-to-end research assistant for planning, executing, and synthesizing qualitative and quantitative user studies.
AweSeed SD-WAN
AweSeed SD-WAN is a comprehensive networking solution that enables users to create high-speed virtual LANs and establish secure cross-site connectivity without requiring a public IP address.
Featured*
