A Multi-layered Methodology for Detecting Astroengineering Technosignatures in Large-scale Transient Surveys and Multi-messenger Astrophysics

The search for technosignatures has historically emphasized narrowband radio emission, directed optical beacons, and quasi-static astroengineering artefacts. A complementary approach is to search for rare transient phenomena that may arise if advanced civilizations manipulate stars, accretion flows, compact objects, or circumstellar media at scale. We propose a formal multi-layered detection framework for identifying such anomalies in large-scale time-domain surveys and follow-up archives. The framework combines physically motivated feature extraction, information-theoretic measures of temporal order and compressibility, anomaly scoring with adaptive isolation forests augmented by expert priors, and multi-messenger consistency tests using gravitational-wave, neutrino, and high-energy electromagnetic data when available. This work was developed and formatted with the assistance of AI tools: Grok (built by xAI), DeepSeek, ChatGPT, and Gemini. Their contributions significantly accelerated the writing, LaTeX formatting, figure generation (pgfplots), stress-testing of the pipeline, and final polishing of the manuscript. The full PDF contains 4 self-contained figures (light curves, latent space, rejection matrix, dynamic stress-test) and is ready for citation and further peer-review. Keywords: technosignatures; transient surveys; anomaly detection; multi-messenger astrophysics; information theory; time-domain astronomy; machine learning; astroengineering; SETI; LSST.