Astrocook v2: LLM-ready foundation for autonomous analysis of quasar spectra

Abstract

Astrocook v2 represents a major architectural redesign of the well-known quasar analysis tool. It is built on an immutable data core, a decoupled API, and a new PySide6 GUI. This system guarantees stability, high performance, and readiness for external integration. We highlight the successful implementation of a robust SciPy-based Voigt Fitter and a powerful Parameter Constraints workflow. This system enables full parameter linking (Value/Thermal modes) and dynamic evaluation. The core strength is reproducible, controlled analysis, supported by the SystemListV2 as the Single Source of Truth for component grouping and the interactive System Inspector for efficient component manipulation. This stable foundation, with its immutable API and fully constrained physical model, is engineered for integration with Large Language Models (LLMs). Astrocook will leverage AI reasoning for guiding complex analysis sequences, advancing toward autonomous pipelines for future spectroscopic surveys.