Observe. Reason. Decide. Learn. Govern.
Vespera is designed around a simple principle: intelligence should be judged against the evidence it actually had, not the future it later discovered.
Markets are not static problems.
Conditions change, relationships break, information arrives unevenly and uncertainty cannot be engineered away. Vespera is not built merely to produce an answer. It is built to determine whether the evidence justifies having one.
Sometimes intelligence is knowing when there is not enough evidence to act.
Start with what actually happened.
Vespera receives governed point-in-time market and public-world evidence. It is not told beforehand which observation must matter.
Separate evidence from interpretation.
The intelligence investigates change, relationships, uncertainty, disagreement and competing explanations without turning plausible stories into facts.
Risk must be justified.
A decision may be long, short, a change in exposure or no trade at all. Action and abstention are evaluated by the same standard: the evidence available at the time.
Outcomes come after decisions.
Decisions are preserved before outcomes become known. Only afterward can experience be used to identify weaknesses, propose improvements and test whether they transfer.
Discovery is not admission.
New knowledge does not become durable merely because it sounds convincing. Proposed lessons must survive controlled testing, counterexamples and transfer checks.
Experience can teach the intelligence. It cannot rewrite the experience.
Vespera preserves the boundary between WORLD(T), DECISION(T), later outcomes and learning. That boundary is central to knowing whether a system actually understood the market at the time.