A local-first semantic memory solution that replaces cloud embedding APIs with Ollama for secure and private data retrieval.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install openclaw-ollama-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 openclaw-ollama-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 OpenClaw Ollama Memory?
This Openclaw Skills extension empowers developers to implement fully private, offline-capable semantic search within their AI agent workflows. By leveraging Ollama and the nomic-embed-text model, users can maintain complete data sovereignty while benefiting from rapid vector embeddings. Unlike standard cloud solutions, this skill removes the need for external API keys from providers like OpenAI or Gemini, making it a cost-effective and secure choice for local development.
The integration is designed for high performance, utilizing the lightweight nomic-embed-text model to provide accurate search results without requiring heavy hardware. By utilizing Openclaw Skills in a local context, users can ensure that their sensitive data never leaves their local environment while still enjoying the benefits of long-term agent memory.
OpenClaw Ollama Memory Use Cases
- Building private long-term memory systems for local AI agents.
- Developing offline-first AI applications using Openclaw Skills.
- Reducing infrastructure costs by replacing paid embedding APIs with local models.
- Ensuring data privacy for enterprise-grade agentic workflows.
How OpenClaw Ollama Memory Works
- The user installs Ollama and pulls the nomic-embed-text embedding model to their local machine.
- The Openclaw Skills configuration is updated to designate Ollama as the primary provider for semantic memory.
- The OpenClaw gateway is restarted to initialize the local embedding engine connection.
- When an agent processes information, it sends text to the local Ollama API to generate vector embeddings.
- These embeddings are stored locally, allowing the agent to perform semantic lookups during future interactions without internet access.
OpenClaw Ollama Memory Setup
First, download Ollama from the official website and pull the required model using the command line:
ollama pull nomic-embed-text
Verify that the service is running correctly:
curl http://127.0.0.1:11434/api/tags
Update your ~/.openclaw/openclaw.json configuration file to include the Ollama provider details under the agent defaults. This ensures that Openclaw Skills can communicate with your local model:
{
"agents": {
"defaults": {
"memorySearch": {
"provider": "ollama",
"model": "nomic-embed-text:latest",
"remote": {
"baseUrl": "http://127.0.0.1:11434"
}
}
}
}
}
Restart the gateway to apply the new settings:
openclaw gateway restart
OpenClaw Ollama Memory Data Schema & Taxonomy
The skill manages memory data using the following configuration schema within the Openclaw Skills framework:
| Parameter | Type | Description |
|---|---|---|
| provider | string | Must be set to ollama to bypass cloud providers |
| model | string | The specific local model used (e.g., nomic-embed-text:latest) |
| baseUrl | string | The local network address where Ollama is listening |
| memorySearch | object | The root configuration block for semantic indexing |
OpenClaw Ollama Memory Advanced Features
- Support for hybrid search combining semantic results with keyword matching.
- Integration with Maximal Marginal Relevance (MMR) for diverse retrieval results.
- Configurable temporal decay to prioritize more recent agent memories.
- Offline-ready execution for high-security environments using Openclaw Skills.
- Multi-path support for indexing multiple local data directories simultaneously.
SKILL.md
Loading
Related Openclaw Skills
Xiaohongshu CTR Optimizer
An AI agent skill that generates high-CTR Xiaohongshu titles using emotional hooks and platform-specific algorithms.
Gold Monitor CN
A comprehensive monitoring tool for the Chinese gold ecosystem, tracking ETFs, spot prices, international futures, and mining stocks.
Gaokao English Vocabulary Frequency Grading System
An interactive, data-driven web application for mastering Gaokao English vocabulary through frequency-based classification and dark-themed UI.
Ghosthand
Ghosthand is a high-performance local HTTP runtime that enables AI agents to observe and control Android devices through structured UI inspection and semantic interaction.
API Consumption Optimizer
A dynamic strategy engine that adjusts API consumption patterns based on remaining quota and reset deadlines to maximize resource utilization.
Paegents Pay
Paegents Pay enables AI agents to securely buy and sell services using bilateral escrow, metered usage agreements, and stablecoin settlements.
Featured*
