Agent

Configuring the Agent

R2R’s Agentic RAG orchestrates multi-step reasoning with Retrieval-Augmented Generation (RAG). The agent is configurable

Learn more about the agent here.

1	[agent]
2	rag_agent_static_prompt = "static_rag_agent"
3	rag_agent_dynamic_prompt = "dynamic_rag_agent"
4	rag_tools = ["search_file_descriptions", "search_file_knowledge", "get_file_content"]
5	research_tools = ["rag", "reasoning", "critique", "python_executor"]