Back to Apps
Wiki-RAG
Supports UIby moodlehq
Experimental RAG system for ingesting MediaWiki sites and providing a high-quality AI interface for knowledge retrieval.
0 stars
0 views
Works in:claude
Exposes:ToolsResourcesPrompts
What it does
Wiki-RAG integrates MediaWiki sites as a knowledge base, using Retrieval-Augmented Generation (RAG) to provide contextually accurate responses. It parses MediaWiki APIs to extract structured data and stores it in a Milvus vector database for hybrid keyword and vector search.
Tools
retrieve: Fetch relevant document chunks from the ingested MediaWiki site.optimise: Refine retrieved documents using popularity (POP) and parent-child (POC) boosting.generate: Produce high-quality responses based on the retrieved wiki context.
Installation
Requires Python 3.12+ and Milvus. Install the package:
pip install -e .
To use as an MCP server, add the wr-mcp executable to your configuration:
{
"mcpServers": {
"wiki-rag": {
"command": "wr-mcp",
"args": []
}
}
}
Supported hosts
- claude
Quick install
pip install -e .Information
- Pricing
- free
- Published
- 5/31/2026
- stars
