Developed from CTCE / Gebit and TBA; (https://doi.org/10.5281/zenodo.17049651)CTGV presents a proof of concept (PoC) of cognition through dynamic geometry: an alternative, auditable, and scalable approach to traditional artificial intelligence systems. Based on topological units (Gebits) of distinct formats:— DECISION MAKER, RESONATOR, and INHIBITOR —, they model complex relationships in a mesh of vector fields, structural interference, and traceable topological propagation. Its architecture is transparent, energy-efficient, and versatile, operating from simple CPUs using graphs and pictograms, allowing its adaptation to multiple computational substrates. Designed for applications that require explainability, robustness, and verification:Anomaly detection, logistics optimization, epidemiological simulation, climate simulations, and critical regulated systems — CTGV is a versatile tool that can be tested, adapted, and applied. Available today for study and development in the repository: (https://github.com/Bear-urso/CTGV-System-V-1.5) The program is a cognitive "mechanism/extension" for governable AI, based on an autonomous framework alternative to blind statistical computing.