Frequently Asked Questions about UBOS MCP Lucene Server

Q: What is the MCP Lucene Server?

A: The MCP Lucene Server is a Java-based implementation of the Model Context Protocol (MCP) that uses Apache Lucene for efficient search and retrieval. It allows AI models to access and interact with external data sources.

Q: What are the key features of the MCP Lucene Server?

A: Key features include MCP compliance, Lucene-powered indexing, a RESTful API, document management (upsert, delete, list), complex query support, filtering, Spring Boot integration, and Dockerization.

Q: What is MCP?

A: MCP (Model Context Protocol) is an open protocol that standardizes how applications provide context to Large Language Models (LLMs).

Q: What are the prerequisites for using the MCP Lucene Server?

A: You need Java 11 or higher, Maven 3.6.0 or higher, and optionally Docker if you plan to use the Docker image.

Q: How do I install the MCP Lucene Server?

A: Clone the repository from GitHub, then build the project using Maven with the command mvn clean install.

Q: How do I run the MCP Lucene Server?

A: You can run it without Docker by executing the jar file java -jar target/mcp-lucene-server-0.0.1-SNAPSHOT.jar or with Docker by building and running the Docker image.

Q: What API endpoints does the MCP Lucene Server provide?

A: The server provides endpoints for checking status (/mcp/v1/status), upserting documents (/mcp/v1/upsert), querying (/mcp/v1/query), deleting (/mcp/v1/delete), and listing documents (/mcp/v1/list).

Q: How do I configure the server?

A: Configure the server using Spring Boot’s application properties, such as server.port and lucene.index.path, in application.properties or application.yml.

Q: What is UBOS and how does the MCP Lucene Server integrate with it?

A: UBOS is a full-stack AI Agent Development Platform. The MCP Lucene Server enhances UBOS by providing a robust search and retrieval mechanism, improving the contextual awareness of AI agents within the UBOS ecosystem.

Q: Can I use the MCP Lucene Server for commercial purposes?

A: Yes, the project is licensed under the Apache 2.0 License, which allows commercial use.

Q: Where is the Lucene index stored?

A: The path to the Lucene index directory can be configured via the lucene.index.path property. If not set, a default location is used, but it’s recommended to configure this to a persistent storage location.

Q: How do I upsert documents?

A: Use a POST request to /mcp/v1/upsert with a JSON body containing an array of documents to be inserted or updated.