Canonical definition
What Is KRYOS V6
KRYOS V6 is a governed framework for structuring decisions where the evidence is incomplete, the stakes are real, and the reasoning has to survive scrutiny afterwards.
KRYOS V6 describes how a complex decision should be framed, evidenced, challenged, and released. It is not a single piece of software and it is not tied to one industry. The same six stages: mission, scope and authority, evidence, contradiction, scenarios, release: apply whether the decision concerns a supplier, a policy option, a grant portfolio, a service redesign, or a security incident.
What distinguishes it from a checklist is the release logic. Most processes end with a recommendation regardless of how strong the support turned out to be. KRYOS V6 treats the strength of support as a gate: weak support produces a narrower output, an escalation, or nothing at all.
Formally, KRYOS V6 can be read as a governed pipeline over a mission object. The mission carries scope predicates and an authority binding. Evidence items are registered with provenance metadata and a fitness assessment scoped to the mission rather than to the source in general.
Contradiction handling is a directed defeater search rather than passive reconciliation: the analysis explicitly enumerates what would falsify the working conclusion. Scenario analysis holds assumption sets as first-class objects so that sensitivity is attributable to specific assumptions.
Release is a classification function over evidential support, contradiction status, and authority coverage, returning proceed, qualify, constrain, escalate, or block. The default under insufficient support is deny, not weak-issue.
Core definition
A governed decision framework: a defined sequence of reasoning stages, each with recorded inputs and a recorded outcome, bound by explicit limits on scope and authority, and terminating in a release classification rather than an unconditional recommendation.
Every term in that sentence is doing work. Governed means the limits are declared before the analysis, not negotiated after it. Framework means it describes structure, not implementation. Release classification means the output is categorised by how much the evidence actually supports.
Why it exists
Serious decisions fail in recognisable ways: the question drifts, the strongest source is the loudest one, disagreement is smoothed into consensus, a single projection is treated as a forecast, and a recommendation is issued because one was expected rather than because it was earned.
KRYOS V6 exists to put a named step in front of each of those failure modes. It does not remove judgement: it makes judgement visible and locatable.
The problem it solves
The core problem is asymmetry between confidence and support. Organisations are good at producing confident output and poor at recording how well supported it was. Six months later nobody can reconstruct what was known at the time.
The framework's contribution is a trail: what the mission was, which sources were used and how they were rated, what contradicted them, which assumptions the conclusion rested on, and why the output was released in the form it was.
The operating principles
Declare before you analyse
Mission, scope, and authority are fixed before evidence is gathered, so scope cannot expand to fit a preferred answer.
Fitness over reputation
A source is rated for this question, not in general.
Search for defeaters
Actively look for what would prove the conclusion wrong.
Assumptions are visible
Scenarios name their assumptions rather than embedding them.
Advice is not authority
Output informs a decision owner; it never substitutes for one.
Fail closed
Weak support produces constraint, escalation, or nothing.
What makes it portable across sectors
The portable parts are the ones that contain no domain content. "Define the mission" and "assess source fitness" mean the same thing to a grants team and to an incident responder; only the sources, the vocabulary, and the authority structure differ.
The non-portable parts are exactly the parts each organisation must supply: what counts as good evidence in your field, who holds authority, which constraints are statutory, and what an acceptable residual risk looks like.
What it does not claim
- It does not claim to produce correct answers, only better-structured and better-recorded ones.
- It does not claim autonomy; nothing is decided or executed without a person.
- It does not claim universal coverage of decision types or domains.
- It does not claim that a modelled scenario is a prediction.
- It does not claim to satisfy any regulatory, clinical, or legal obligation.
- It does not claim measured improvement in decision outcomes.
Apply this to your sector
See how the same stages read in business, government, NGO, healthcare, legal, education, operations, research, and cybersecurity settings.