Concepts

WORLDLINE has seven ideas. Each one has its own page, and each page builds on the ones before it.

Concept	One line
Worlds	A deterministic simulation where truth, what each role has received, and what the evaluator concludes are kept apart.
Interventions	Changes to how the environment delivers events (delay, drop, duplicate, reorder, corrupt, revoke, clock skew), which branch one world into many.
Invariants	Guarantees that are checked in every world and return PASS, FAIL with a witness, or UNKNOWN with the missing signal.
Minimization	Delta debugging that reduces a failing set of interventions to a 1-minimal cause and checks that result independently.
Failure genome	The cause, abstracted from event ids to structure, so that it can challenge a different scenario.
Evidence kinds	Every claim is typed by how it was obtained: PROVED_WITHIN_MODEL, EMPIRICALLY_VALIDATED, ATTESTED, FAILED, or UNKNOWN.
The kernel	The one component that recomputes agent claims, attests what matches, seals the evidence, and authorizes a release for one exact commit.

flowchart LR
  W[Worlds] --> I[Interventions] --> V[Invariants] --> M[Minimization] --> G[Genome]
  V & M & G --> E[Evidence kinds] --> K[Kernel]