CAUSENTIA v2.2: Sovereign Crisis Early Warning System — Research Papers, Platform Source Code, Python SDK & Reproducibility Package

CAUSENTIA — Sovereign Crisis Early Warning System (v2.2) CAUSENTIA is an open-source sovereign crisis early warning system that monitors 80 countries in real time using 25+ quantitative indicators sourced from the World Bank, the Federal Reserve (FRED), and GDELT. The platform computes the Collapse Index (CI), an original composite metric (0–100) that decomposes sovereign risk into three structural dimensions: Stress, Absorption Capacity, and Institutional Resilience. Backtested against 15 major sovereign crisis episodes (2010–2025), the CI achieves 87% recall, 86% F1-score, and an average lead time of 10.1 months, outperforming established benchmarks including the KLR Signal Approach and the IMF Vulnerability Index.All benchmarks were reimplemented under identical data availability and evaluation windows to ensure methodological comparability. This record contains Five research papers covering the full methodology, systemic risk indices (CausalEntropy Index & Fracture Index), contagion network analysis, and software documentation Platform source code, including a Python FastAPI backend, JavaScript frontend, and six RESTful API endpoints Python SDK for programmatic access to the CAUSENTIA API Reproducibility package with build system, SHA256 checksums, and full instructions to regenerate all papers from source Key innovations Country Deep Dive (state-level temporal diagnostics) Regional Risk Heatmap (cross-region systemic exposure) Scenario Builder (6 predefined scenarios + 8 custom policy shock sliders) Monte Carlo Simulation (10,000 scenarios per country with forward-looking stress propagation) Contagion Network (85+ trade and financial transmission links) AI Analyst Chat (explainable, data-grounded system diagnostics) Early Warning System (four-tier alert structure) Automated PDF Report Generator for institutional use Resources Live system: https://causentia.org Source code: https://github.com/causentia/causentia Intended use This release is intended to support academic research, policy analysis, and institutional early-warning experimentation. Disclaimer All outputs are probabilistic monitoring signals derived exclusively from publicly available data. They do not constitute financial advice, investment recommendations, sovereign credit ratings, or political assessments. The Collapse Index is not a credit rating as defined under EU Regulation (EC) No 1060/2009 or the U.S. Credit Rating Agency Reform Act of 2006.Each paper includes a comprehensive Legal Disclaimer and Terms of Use (11 clauses). Please refer to the individual papers for full terms. License Software: MIT License Text & Figures: CC BY 4.0