viewing as CISO, administrator CRO Actuary, model owner Data lead, register owner Validator Compliance lead Internal audit Business line head, requester Procurement lead The prototype switches the signed-in role. Only the named role sees a decision's trigger.
Conformance self-tests (self-declared) Prototype states
Results Run did not complete Never run
Self-declared. GovernanceOS implements the Defensible AI Framework; the registry is a specification, not a certification scheme.
Framework Self-test Result, run 2026-10-16 Run log
Five-Gate framework (REG-06) one approver per gate passed Run log Five-Gate framework (REG-06) G2 approver is a different person from G3 passed Run log MESA model risk management (REG-02) every validation report lists what was not tested failed: pricing model v3 Run log Exit covenant export works under a simulated expired license passed Run log
Learn more about REG numbers REG numbers are entries in the Defensible AI Framework Registry. MESA is the readiness and model risk method; Five-Gate is the G1 to G5 spine. Each self-test links to its run log below.
Run logs
Run log: one approver per gate nebula-mutual/conformance/five-gate-approver
Framework Five-Gate framework (REG-06) typed
Run 2026-10-16 02:00 CT , nightlyproduced
Result passed produced
Output every record checked; no exception. produced
Run hash rrun-five-ga… show full hash rrun-five-gate-approvera9edbc74c7346ddabc4ccfc6b4fc1fc790e06acfe produced Run log: G2 approver is a different person from G3 nebula-mutual/conformance/five-gate-g2g3
Framework Five-Gate framework (REG-06) typed
Run 2026-10-16 02:00 CT , nightlyproduced
Result passed produced
Output every record checked; no exception. produced
Run hash rrun-five-ga… show full hash rrun-five-gate-g2g385e0f9af4e7429d6cdea322be040077c50c607448ef93 produced Run log: every validation report lists what was not tested nebula-mutual/conformance/mrm-unexamined
Framework MESA model risk management (REG-02) typed
Run 2026-10-16 02:00 CT , nightlyproduced
Result failed produced
Output 3 reports checked; 2 list what was not tested; pricing model v3's G2 report has no unexamined set. produced
Run hash 77c1d3290765… show full hash 77c1d3290765c057729ba1ae99e4bdfbbff8bc83b7ac2be0b7fa447c4dd8da98 produced Run log: export works under a simulated expired license nebula-mutual/conformance/exit-expired
Framework Exit covenant typed
Run 2026-10-16 02:00 CT , nightlyproduced
Result passed produced
Output every record checked; no exception. produced
Run hash rrun-exit-ex… show full hash rrun-exit-expired5947a3b2b48314a1f746fb7cbbb408cb78c5ce08fb15f57 produced
No self-test has run yet The first run starts at 02:00 CT tonight, after the regime profile is set. Results appear here with their run logs.
Defensible AI Platform
Nebula Mutual
