Private Knowledge Base FAQs

Question 1

Which LLMs and AI clients are compatible with Private Knowledge Base?

Accepted Answer

It's compatible with any LLM and AI client that supports the MCP protocol, including Claude Desktop, Cursor/VS Code, Claude Code (CLI), and local LLMs like Ollama when used with tools like Continue.dev.

Question 2

What is Private Knowledge Base and how does it ensure data privacy?

Accepted Answer

Private Knowledge Base is an MCP server that provides any LLM (cloud or local) with a RAG-powered knowledge base and web search. It guarantees privacy by performing all document chunking, embedding, and indexing locally on your machine, ensuring your sensitive data never leaves your environment.

Question 3

Is Docker required for setting up Private Knowledge Base?

Accepted Answer

No, Docker is not required for the default local file-based storage mode. The server runs out-of-the-box with zero-ceremony setup. Docker is an optional choice for scaling up with a Qdrant server instance.

Question 4

How do I add documents to my Private Knowledge Base?

Accepted Answer

Simply drop .md, .txt, or .pdf files into your designated 'data/' folder. The server automatically loads, chunks, embeds, and indexes them on startup. You can also ingest new documents at runtime using the `ingest_document` tool via your AI client.

Question 5

Does this tool provide web search capabilities for LLMs?

Accepted Answer

Yes, Private Knowledge Base includes integrated live web search via Firecrawl, alongside semantic search over your local documents. This is particularly valuable for local LLMs (e.g., Ollama) that typically lack built-in web access or other tools.

Private Knowledge Base

Private Knowledge Base

主要功能

使用案例

主要功能

使用案例