A high-performance Model Context Protocol server providing Retrieval-Augmented Generation (RAG) and semantic search tools for AI agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install openviking-mcp
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 openviking-mcp 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 OpenViking MCP Server?
The OpenViking MCP server is a powerful integration designed to bring professional-grade Retrieval-Augmented Generation capabilities to AI environments. By leveraging the Model Context Protocol, this skill allows users to expose their private documentation and data repositories to Claude and other MCP-compatible clients. This makes it a standout choice among Openclaw Skills for developers needing precise document Q&A. This server handles the heavy lifting of semantic indexing and retrieval, transforming static files into an interactive knowledge base. It provides a bridge between local data and large language models, ensuring that AI responses are grounded in specific, user-provided context.
OpenViking MCP Server Use Cases
- Powering document-heavy Q&A sessions within Claude Desktop or CLI.
- Performing semantic searches across local directories, PDFs, and remote URLs.
- Building a centralized knowledge base for specialized Openclaw Skills applications.
- Automating the ingestion of web resources into a searchable RAG pipeline.
How OpenViking MCP Server Works
- The user installs the OpenViking repository and its Python dependencies using the uv package manager.
- Configuration of API tokens for LLM and embedding providers is handled via a local configuration file.
- The server starts as a background process, exposing tools like query and search via HTTP or stdio transport.
- MCP-compatible clients connect to the server to access the RAG pipeline.
- When a query is received, the system performs a semantic search to find relevant context before generating a response.
OpenViking MCP Server Setup
Clone the repository and install dependencies using uv:
git clone https://github.com/ZaynJarvis/openviking.git
cd openviking/examples/mcp-query
uv sync
Configure your credentials by copying the example config:
cp ov.conf.example ov.conf
Edit the ov.conf file to include your OpenAI API key for vlm.token and embedding.token. Once configured, start the server:
uv run server.py
Finally, connect your Claude Desktop or CLI to the local endpoint, typically http://localhost:8000/mcp.
OpenViking MCP Server Data Schema & Taxonomy
The OpenViking MCP server organizes its data through a structured configuration and a dedicated data directory. It supports the following components:
| Component | Description |
|---|---|
ov.conf |
A JSON-based configuration file defining LLM providers, model types, and API tokens. |
| Data Directory | A local path (defaulting to ./data) where indexed embeddings and persistent storage are maintained. |
| Resource Types | Supports ingestion of local files, entire directories, and remote web URLs via the add_resource tool. |
| Tool Schema | Exposes specific JSON schemas for query, search, and add_resource to ensure client-server compatibility. |
OpenViking MCP Server Advanced Features
- Support for both streamable-http and stdio transport protocols to fit various architecture needs.
- Flexible configuration via environment variables such as OV_CONFIG, OV_DATA, and OV_PORT.
- Dynamic resource indexing allowing for real-time updates to the knowledge base without server restarts.
- Multi-tool interface providing separate endpoints for raw semantic search and full RAG-driven answer generation.
- Built-in troubleshooting hooks for managing port conflicts and authentication validation.
SKILL.md
Loading
Related Openclaw Skills
Email Processor
A robust Gmail automation tool that triages your inbox by archiving low-priority marketing emails while surfacing critical communications.
YouTube Watcher
Fetch and process YouTube video transcripts for instant summarization and content analysis.
YouTube Watcher
An essential skill for fetching and reading YouTube video transcripts to enable AI-driven summarization and information retrieval.
x-trends
A powerful CLI-driven skill to search, filter, and analyze real-time trending topics and hashtags on X (Twitter) across global locations.
OpenViking
A next-generation RAG system that uses a filesystem paradigm to manage AI agent memories, resources, and skills.
Sophie Optimizer
An automated context management utility that monitors token usage, snapshots memory, and resets sessions to maintain peak AI performance.
Featured*
