Optical QR reels
Camera-readable chunks carry the signed object itself. Physical Mac-to-Mac optical acquisition has been verified at 30/30 chunks; QR is transport, not semantic reinterpretation.
Braid moves signed AI state across machines without forcing every exchange through a cloud intermediary. A sender can bind explicit semantic material to its local model provenance; a receiver verifies that evidence and, when model spaces differ, re-embeds the meaning locally in its own native space.
Standalone macOS app in a drag-to-Applications DMG. Local research build; see the Gatekeeper note below before launch.
The transport layer carries the exact signed .brad bytes. It does not reinterpret, re-sign, or silently admit them. The receiving machine quarantines the object, evaluates Phase A evidence, and only marks acceptance after its own Phase B transaction commits finality.
Camera-readable chunks carry the signed object itself. Physical Mac-to-Mac optical acquisition has been verified at 30/30 chunks; QR is transport, not semantic reinterpretation.
Signed state can move directly between private-network peers while the receiver retains quarantine, replay protection, validation, and finality authority.
Different embedding-model digests trigger receiver-local re-embedding of explicit signed semantic material. No opaque vector-to-text guessing and no invented semantic bridge.
This downloadable v1.5.2 DMG is not yet distributed through Apple's Developer ID notarization pipeline. Gatekeeper can therefore show “Braid was blocked to protect your Mac.” You do not need to disable macOS security.
Use the one-app exception built into macOS. The wording can vary slightly by macOS version.
A local-first research client for signed state transport and explicit cross-model semantic handoff. Evidence may travel; authority remains local.
.brad frameAI systems should be able to exchange useful state without surrendering identity, provenance, or operator control.
Braid is being built as an open coordination primitive for local and sovereign AI systems: compact signed objects, receiver-owned policy, explicit semantic continuity, and transport flexibility from a camera lens to a private network. The design rule stays simple: evidence may travel; authority remains local.