Iron-Dome-but-the-brain
A precise framing of where SZL sits in a layered-defense picture: the brain, not the missile. SZL builds the governed decision layer — the part that decides, under proof, what a recommendation is — and explicitly does not build effectors.
The distinction
An interceptor system has two halves: a decision half (detect, classify, prioritise, decide whether and how to respond) and an effector half (the kinetic action). SZL is entirely in the decision half, and only the governed part of it:
| Layer | Who owns it | SZL? |
|---|---|---|
| Sensing / detection | sensors, radar, broadcast feeds | partial — killinchu decodes broadcast self-ID |
| Classification / track | tracking systems | partial — killinchu track + swarm topology |
| Governed decision | the brain | yes — this is SZL |
| Effector / engagement | weapons systems | no — out of scope by design |
What "the brain" actually does
- Bounds the action space — the Bekenstein / geofence invariant (INV-4) makes the set of considered actions finite and physically feasible (the Killinchu-bridge
G(a)factor). - Gates every candidate — the 13-axis Λ-gate admits a recommendation only if all 13 axes clear their floors (conjunctive AND).
- Halts safely — the HUKLLA soft-halt drives utility to zero on any fired tripwire; T10 (STOP) is absorbing (INV-1).
- Receipts everything — every decision emits a Khipu receipt, so the why is replayable (INV-3).
The output is an auditable recommendation with a proof chain — not a fire command.
Why this framing matters
It is the honest, defensible scope. SZL's value is that a human reviewer can replay the brain's reasoning with cosign and Lean and see exactly which axis admitted or rejected an action. That is the Warhacker thesis applied to a kinetic domain: delivery + verification + governance, never the weapon.
Doctrine v11 scope discipline
No autonomous lethal action. A human remains in the loop for any kinetic outcome. SZL produces the governed decision and its receipt; it does not actuate.