Lucenia Assistant Toolkit

Introduced 0.5.0

The Lucenia Assistant Toolkit provides the building blocks for AI-powered applications that run inside the Lucenia cluster. The toolkit includes the following elements:

• Agents and tools: Agents interface with a large language model (LLM) and execute high-level tasks, such as summarization or retrieval-augmented generation (RAG). Lucenia provides three agent types: flow, conversational_flow, and conversational. Each agent's high-level task consists of low-level tasks called tools, which can be reused across agents. Lucenia ships eight built-in tools, including SearchIndexTool, CatIndexTool, ListIndexTool, IndexMappingTool, MLModelTool, AgentTool, ConnectorTool, and VisualizationTool.
• MCP tools: An experimental Model Context Protocol (MCP) server lets external clients and agents discover and invoke Lucenia tools over stateless HTTP.

Enabling the agent framework

Enable the agent framework and retrieval-augmented generation (RAG) by configuring the following settings:

plugins.ml_commons.agent_framework_enabled: true
plugins.ml_commons.rag_pipeline_feature_enabled: true

Next steps

• For an overview of agents and tools, see Agents and tools.
• For a step-by-step tutorial, see Agents and tools tutorial.
• For using MCP tools, see Using MCP tools.