Cardea Agentic Search
Provides agentic search capabilities with support for vector search using Qdrant, full-text search using TiDB, or both combined within a Model Context Protocol (MCP) server.
About
Cardea Agentic Search delivers an intelligent, comprehensive search experience by integrating semantic vector search with Qdrant and traditional full-text keyword search via TiDB. Operating as a Model Context Protocol (MCP) server, it leverages external chat services for dynamic keyword extraction and embedding services for generating vector representations, ensuring highly relevant and nuanced search results. Users can flexibly configure search modes, combining the strengths of both approaches to power advanced retrieval-augmented generation (RAG) applications and other AI-driven information systems.
Key Features
- Vector search using Qdrant and embedding services
- Keyword search with TiDB and intelligent keyword extraction
- Combined search for comprehensive results
- Flexible configuration of search modes
- Customizable keyword extraction prompts
- 2 GitHub stars
Use Cases
- Integrating semantic and keyword search into applications
- Developing intelligent document retrieval systems
- Building advanced AI-powered search engines
