failure-prediction
A unified framework for tabular probabilistic regression, time-to-event prediction, and probability distributions in python
Predict LLM-agent-run failures in real time — loop/cascade/drift predictors, RAG grounding guard, failure-memory RAG, and a trace-replay backtest. ProactiveGuard re-targeted at agents.
Python survival analysis/survival curves, with a scikit-learn-like api
