Calibrated incident root cause inference for TypeScript: declarative Bayesian networks with exact and approximate inference and the do-operator, zero runtime dependencies.
nodejs typescript mcp incident-response bayesian-network edge calibration openai bayesian zero-dependency sre observability causal-inference diagnosis belief-propagation root-cause-analysis llm anthropic probabilistic-graphical-model do-operator
-
Updated
Jun 23, 2026 - TypeScript