Supreme Computation is backed by public implementation, tests and evidence so technical people can challenge what the system actually does.
Live governance states, 6/6 adversarial execution tests, a successful automated GitHub self-test, and the public implementation in one challengeable proof surface.
Inspect the reference architecture directly.
Actions can proceed only after required conditions qualify.
Missing, stale, contradictory or incomplete evidence can stop release.
Actions crossing policy, authority or boundary requirements can be refused.
Execution can preserve evidence connecting decisions with what actually happened.
New information can enter without silently erasing previously proven state.