API reference

Every public class and function, grouped by layer. The top-level package re-exports all of it, so from hiphopsllm import X works for anything listed here.

Layer	What it does
pipeline	AgenticReliabilityStudy — the whole chain behind one object
architecture	LangGraph → components, ports, connections; loop elimination
faulttree	Failure logic, synthesis, cut sets, quantification, FMEA, exports
reliability	Operational profiles, HIP-LLM, evidence calibration
bayes	CPTs, Bayesian networks, learned tables, drawing
viz	Matplotlib rendering of trees, architectures and rankings
report	SafetyReport and analyse_langgraph
io	Bundled example architectures and outcome data

The 30-second version

from hiphopsllm import AgenticReliabilityStudy

study = AgenticReliabilityStudy(graph, globals_ns=globals())
study.observe(outcomes, profile={"short": 0.3, "long": 0.7}).run()

study.summary()                  # everything, printable
study.hazard_probability("H2")   # an Envelope
study.bayesnet("H2")             # a BayesianNetwork
study.save("artifacts/")