Why a solution without a test is not yet a solution
A contradiction card with a "get both" move looks like a win. But without validation it is a hypothesis in nice packaging. The Super Oracle passport states the boundary directly: the agent does not promise novelty, does not run a patent search, and the proposals require engineering, product and economic evaluation by the relevant specialists. The only way to turn a move into knowledge is a test.
Three fields for every move
The "Validation" stage of the agent's process sets three fields for every direction:
- Expected effect. What exactly must change and how it will be measured — before the test starts, not after.
- Risk. What can go wrong if the move is wrong, and what damage it will bring.
- Cheap test. A minimal check that rules a move out in days, not months of development.
The test must be cheaper than the implementation: if checking the hypothesis requires building the whole product — that is not a test, that is a gamble.
Prioritization by RICE
When there are several moves, the argument about which one to check first is settled by the RICE formula from Oracle's arsenal: reach × impact × confidence ÷ effort. Every direction gets a number, and the team gets a ranked list and the first three steps instead of arguments and intuition. In the final artifact of the session this is a priority table: what to do first, what not to do at all.
The link with Pre-mortem
The risk from a move card is useful to run through Pre-mortem — the method of dissecting a failure in advance: the team imagines that the project has already failed and names the causes. For every test this gives a "do NOT do" checklist and a prevention plan for the main risks. This way the test checks not only the effect, but also protects from the most expensive error scenario.
Boundaries of honesty
It is important what this stage does not promise: it does not promise novelty (no patent search is done), gives no legal opinion and does not replace the evaluation of specialists. The stage does something else — it turns every proposal into a verifiable form where a failed test costs little and happens early. How the set of moves is expanded before validation — in the article "Solution-space expansion".
Questions and answers
What counts as a cheap test?
A check that is several times cheaper than the implementation and gives a result in days: a prototype, a pilot on one process, a survey about past behavior, a manual run of the method on one example.
What to do if the test fails?
Go back to the contradiction map and take the next direction — that is the point of a set of fundamentally different moves: the failure of one does not kill the task.
Who decides on the test results?
The team and the relevant specialists: Oracle artifacts are material for engineering, product and economic evaluation, not a directive.
What does the test for a move include?
The expected effect, the risk and a cheap way to check.
How to prioritize the tests?
By RICE: reach, impact, confidence and effort — the artifact provides material for such an evaluation.
What is a cheap test?
A minimal check of a hypothesis without the full implementation: a mock-up, a pilot, an observation.
Does Oracle ensure novelty?
No: patent search and legal expertise remain with the relevant specialists.
What to do with the risks from the map?
Include them in the validation plan: every risk is linked to a specific move.
Who interprets the test results?
The team and the relevant specialists; the artifacts are material for evaluation, not a directive.
Can the method be used for a service?
Yes: contradictions exist in processes, services and business models.