Aerospace Instrument-Making Annotation << Back Prospects for Using Intelligent Agents in Space Exploration Tyugashev A.A., Koshelev A.Yu., Tkachenko I.S. The presented article discusses prospects for using intelligent agents based on large language models (LLMs) in space exploration. Key application areas are outlined: AI assistants for engineering design (including RAG, CAD workflows, and generative design), reliability analysis and predictive maintenance using telemetry and digital twins, intelligent robotics for highly autonomous missions, and onboard AI assistants for crews and personnel training. Special attention is paid to applying LLM-based agents to onboard software engineering and requirements engineering for safety-critical systems; the paper proposes the principle “AI proposes → human approves → mathematics proves,” emphasizing requirements clarifi cation and formalization, traceability management, test generation, and integration with formal verifi cation methods. The paper also analyzes limitations and risks (hallucinations, non-determinism, confi dentiality constraints, and computing requirements) and describes mitigation approaches such as strict RAG, on-premise deployment, and mandatory human expert review. Keywords: space exploration; artificial intelligence; intelligent agents; large language models; LLM; Retrieval-Augmented Generation (RAG); CAD; generative design; digital twin; predictive maintenance; telemetry; space robotics; onboard software; requirements engineering; formal verification; safety; confidentiality. DOI: 10.25791/aviakosmos.2.2026.1538 Pp. 39-48.