This paper presents a feasibility framework for detecting technological activity through information thermodynamics. It compares radiative entropy export, deliberately structured emission, persistent stellar modifications, and compact-object signatures. Under the assumptions of large-scale irreversible computation and thermal waste disposal, the framework relates a star's radiative entropy excess to the captured luminosity and radiator temperature, and expresses an upper bound on bit erasure. It also considers coherence and model-based complexity measures, including algebraic redundancy and repeated complex profiles, as proposed tests of structured modulation. Natural variability, instrumental contamination, incomplete null models, and the distinction between anomaly detection and artificial-origin claims are central to the proposed assessment procedure. Compact-object channels are evaluated through emission budgets, radius–temperature discrimination, and population statistics, while composition and dynamics are considered as possible long-lived archaeological tracers. The framework prioritizes waste-heat searches followed by independent observations and contamination vetoes. Its applicability is limited by non-thermal energy disposal, reversible computation, cold radiators, and uncertainty in natural baselines. The proposed detection tests and safeguards are presented as a framework for future investigation, not as a completed observational search or a confirmed detection of extraterrestrial technology.