Lilbee FAQs

Question 1

How can AI agents integrate with Lilbee?

Accepted Answer

AI agents can integrate with Lilbee using its built-in MCP server, which offers tools like `lilbee_search`, `lilbee_ask`, and `lilbee_status`. Alternatively, agents can use Lilbee's JSON CLI for structured output, enabling flexible scripting.

Question 2

Is Ollama always necessary when using Lilbee?

Accepted Answer

Ollama is required for embedding documents during the initial indexing ('sync') and for generating answers in interactive chat ('ask') sessions. However, the `lilbee search` command, often used by agents, works without Ollama once documents are indexed, returning pre-computed embeddings.

Question 3

Does Lilbee require cloud APIs or external databases?

Accepted Answer

No, Lilbee is designed to be fully local-first. It runs entirely on your machine, using Python, Ollama, and LanceDB, without relying on any cloud APIs, external databases, or Docker containers for core functionality.

Question 4

What types of documents and code can Lilbee index?

Accepted Answer

Lilbee supports a wide array of formats, including PDFs, Office documents, eBooks, images (with optional OCR via Tesseract), CSV/TSV, plain text, and over 150 code languages. Code files benefit from AST-aware chunking via Tree-sitter for improved context.

Question 5

What is Lilbee and what does it do?

Accepted Answer

Lilbee is a local Retrieval Augmented Generation (RAG) CLI and MCP server. It indexes your documents and code on your machine, allowing you to ask questions via an interactive terminal chat or integrate its knowledge base with AI agent tools.

Lilbee

Lilbee

Características Principales

Casos de Uso

Características Principales

Casos de Uso