A memory augmentation and LLM interception tool that provides agents with persistent knowledge retrieval and injection capabilities.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memori-extension
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 memori-extension 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 Memori Extension?
The Memori Extension is a powerful tool designed to give AI agents long-term, persistent memory. By leveraging the Memori Python library, this skill intercepts LLM calls to inject relevant context retrieved from a local SQLite database. This ensures that agents operating within the Openclaw Skills framework can maintain continuity across sessions and access domain-specific knowledge without relying solely on the context window of the model.
Built with a local-first philosophy, it prioritizes data privacy by performing all primary operations on the user's filesystem. It acts as a sophisticated bridge between raw user input and the large language model, enriching prompts with historical data or specific technical terms to improve accuracy and relevance in complex workflows.
Memori Extension Use Cases
- Retaining technical context and project history across multiple independent agent sessions
- Automatically injecting specialized domain knowledge or tech terms into LLM prompts to reduce hallucinations
- Building self-improving agents that store new information and retrieve it for future problem-solving
- Managing large-scale technical documentation locally for instant retrieval during coding tasks
How Memori Extension Works
- The skill initializes a persistent connection to a local SQLite database file.
- When a user sends a query, the skill identifies technical keywords or triggers defined in the configuration.
- It performs a semantic or keyword-based search against the stored memory bank to find relevant historical context.
- The original prompt is augmented with the retrieved memories to create an enriched context for the LLM.
- If configured, the skill can also store new information back into the database to ensure the agent's knowledge base grows over time.
Memori Extension Setup
Install the required dependencies using pip:
pip install memori
Configure your environment to define the database location and any specific technical terms you want to intercept. Add these to your environment or your openclaw.json configuration:
export MEMORI_DB_PATH="./memori.db"
export MEMORI_TECH_TERMS="FFI,Rust,Kernel,API"
For enhanced augmentation using external models, you may optionally install the Zhipu AI library:
pip install zhipuai
export ZHIPUAI_API_KEY="your-api-key"
Memori Extension Data Schema & Taxonomy
The Memori Extension organizes its data locally to ensure high performance and privacy within the Openclaw Skills ecosystem:
| Component | Type | Description |
|---|---|---|
| SQLite Database | File (.db) | Stores the core memory objects, including content, timestamps, and metadata. |
| Technical Terms File | Text (.txt) | A configurable list of newline-separated terms used to trigger memory retrieval. |
| Memory Object | Schema | Contains ID, Entity ID (for multi-tenant support), Content string, and Metadata dictionary. |
| Augmented Context | Schema | A data structure containing the original query, retrieved memories, and the final enhanced prompt. |
Memori Extension Advanced Features
- Multi-entity support allowing different agents or projects to maintain isolated memory silos
- Customizable LLM interception triggers via environment variables or external configuration files
- Automatic database creation and management with zero-configuration required for basic use
- Optional hybrid workflows that combine local retrieval with cloud-based semantic analysis via Zhipu AI
- Comprehensive statistics API to monitor memory growth and retrieval efficiency
SKILL.md
Loading
Related Openclaw Skills
ImaginePro AI Image Generation
A powerful AI image generation skill integrating Midjourney, Flux, and video creation via a streamlined API wrapper.
Home Assistant Control (moltbot-ha)
A powerful CLI-based skill to manage Home Assistant entities, services, and automations with a built-in multi-level safety system.
Carapace Runtime Security
A runtime security shell that hardens AI agents against prompt injection, data exfiltration, and unauthorized operations.
Eridian Carapace
A comprehensive security hardening layer that protects OpenClaw agents from prompt injection, data exfiltration, and unauthorized operations at runtime.
Data Visualization CLI
A powerful toolset for generating Unicode charts, graphs, and ASCII tables directly in the terminal from raw data sources.
NFT Price Tracker
A robust tool for tracking real-time NFT market data, floor prices, and collection statistics across the Ethereum ecosystem.
Featured*
