skills.rest
dingdyan
dingdyanCOMMUNITY
Explore all skills by this author →
GitHub

openviking

Serve RAG and semantic search over documents via an OpenViking MCP interface.

Updated Apr 5, 2026
Search Tags:#semantic search#rag#vector database#knowledge base#mcp#document q&a#resource ingestion
One-click install
npx skills add https://github.com/dingdyan/openclaw-workspace-v2 --skill openviking-dingdyan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: openviking
Source: https://github.com/dingdyan/openclaw-workspace-v2/tree/main/skills/openviking
Command: npx skills add https://github.com/dingdyan/openclaw-workspace-v2 --skill openviking-dingdyan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Openviking solves the problem of turning scattered files, URLs, and knowledge into retrievable context for AI Q&A without relying on flat, single-shot vector storage.

Core Features & Use Cases

  • Filesystem-based RAG memory: Manage context as URI-addressable resources rather than only embedding blobs.
  • Tiered context retrieval (L0/L1/L2): Load progressively more detailed content on demand for more accurate answers.
  • MCP server tools: Perform end-to-end RAG querying, semantic search, and add resources (files and URLs) into the knowledge base.
  • Use cases: document Q&A, knowledge management, AI agent memory, file/URL discovery, and semantic retrieval from PDFs and other documents.

Quick Start

Start the MCP server by running the provided init script (one-time) if needed, then launch the example server and connect your client with claude mcp add --transport http openviking http://localhost:2033/mcp.

Frequently Asked Questions about openviking

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up semantic search over a document knowledge base for AI agents?▼

You can achieve semantic search over a document knowledge base by serving an OpenViking Context Database MCP interface, which requires configuring API keys in ov.conf and launching a local MCP server to handle vector-backed retrieval queries.

Can I add new files and URLs into a RAG knowledge base during an agent workflow?▼

Yes, you can add new files and URLs into a RAG knowledge base during an agent workflow by using the MCP add_resource tool, allowing the vector-backed memory to ingest and discover new content reliably across ongoing processes.

Does the OpenViking MCP server support document Q&A from PDFs and other files?▼

The OpenViking MCP server supports document Q&A from PDFs and other files by using the query and search MCP tools, retrieving progressively detailed tiered context (L0/L1/L2) to provide accurate answers from ingested URI-addressable resources.

What's the best way to manage RAG memory without relying on flat single-shot vector storage?▼

The best way to manage RAG memory without flat single-shot vector storage is treating context as URI-addressable resources within an OpenViking Context Database, enabling tiered context retrieval on demand for more accurate AI Q&A responses.

How do I connect my client to the OpenViking MCP server for document retrieval?▼

To connect your client for document retrieval, start the provided init script and example server, then run the command 'claude mcp add --transport http openviking http://localhost:2033/mcp' to link your client to the local OpenViking MCP server.

Related Skills

Explore More Data & Analytics Skills→
MaosixianMaosixian

openviking

Manage and search structured knowledge with RAG and semantic search via MCP.

Community
Advanced
vasilyu1983vasilyu1983
73

ai-rag

Design and implement RAG retrieval systems with hybrid search and reranking.

Community
Advanced
JamkrisJamkris
86

mcp-server-patterns

Automate MCP server construction with reusable Node/TypeScript SDK patterns.

Community
Advanced
rukikotoorukikotoo
6

rag

Query the econometrics RAG knowledge base to retrieve relevant document chunks.

Community
Intermediate