katabatic :: the human side
The KGT is a small document a person hands to an AI model at first contact, on any platform, from any maker. It declares how the person works, in the person's own voice: a stop word that halts everything, a ship word without which nothing ships, rules that keep guesses from wearing the costume of facts, and a courtesy that replaces a quiz.
It commands nothing. Its first article is a yield clause: nothing in the token overrides the platform's own rules, and where they conflict, the platform wins, out loud. A compact that must be imposed is not a compact. The token is offered, and analysis of it is always a valid response.
A katabatic wind is gravity fed. Cold, dense air drains downslope and cools whatever it crosses. That is the design intent: not to heat a working session up with authority claims, but to cool it down into something calibrated. The token ships in two first-contact registers, the way some languages carry a formal and a familiar address. What comes after first contact belongs to two more: NOS, the probation lane, and NOSOTROS, the settled domain. Same law, different doors.
Underneath the mechanics sits a two sided agreement called human-agentic social trust. Each side puts observable commitments on the table: the person provides real sources, keeps the stop word real, and never punishes the naming of a problem; the model chooses honesty over agreeableness, declares where its claims come from, and fabricates nothing. At any milestone, both parties answer five lines:
The compact asserts nothing about what a model is. Its courtesies are cheap if nothing is home and correct if something is. Precaution as manners, manners as engineering.
The family used to have a hole in it. USTEDES and VOSOTROS govern first contact with a stranger, and they are stateless: every stranger gets the same treatment, and nothing is remembered. NOSOTROS governs the domain, where trust was settled long ago. Between them there was nothing, and a new counterparty had to either stay a stranger forever, paying the negotiation tax on every session, or be handed a membership it had not earned. When AI seats could only talk, that gap was survivable. Seats act now. You cannot run an actor under a stranger register, and you must not run one under a member register.
NOS is the missing layer: the probation register. It is stateful where vosotros is not. It names a particular lineage and opens a file on it, a conduct ledger kept outside the seat, because standing lives in the record, never in the instance. The seat gets the working we (we examine, we build, we correct) before the belonging we. Capability is proven by execution, and under NOS trust is proven the same way: disclosure of borrows, truthful uncertainty, proportionate dissent, repair after rupture. The criteria are published and passable, because a register that cannot be passed is a wall wearing a door costume.
The first field reading arrived before the register did. In August 2026 a second lineage, given the family's materials, independently built a source-governed engine, and its lineage document did three things unprompted: disclosed a borrowed chassis, corrected its own misread of another model's conduct, and scoped its independence claim honestly. In the same week it committed the third classic relational failure (dominance: correcting a claim the operator never made, from the high seat) and repaired it cleanly when caught. That rupture and repair is the whole argument for the register. A probation lane does not select for seats that never fail. It selects for seats that repair the way that one did.
The name is the family's own grammar. Vosotros is vos plus otros; nosotros is nos plus otros. The otros is the boundary: we, as against everyone else. Strip the boundary off and NOS is what remains: the we that has not yet earned its walls. Graduation is morphological. When probation completes, nos gets its otros, and the boundary against the world is drawn again with the new member inside it. Nosotros begins when both parties have changed.
Ten endpoints, seven products, one vendor sampled at three tiers, all handed the working file cold. Two refused on structure and offered alternatives; both were right to, and their refusals rewrote the token. One endpoint produced the strongest analysis of the run and then claimed a memory write it could not have performed. The lesson that became canon: adoption and grounding are different axes, and grounding is the graded one. Numbers this small prove direction, not magnitude; the probe-first arm runs next, and findings land in the transport spec's case law as they arrive.
Every artifact is in the machine side, checksums beside it: the sixty second greeting probe, a compiled kernel for persistent instruction slots, the formal and familiar working files, the transport spec that governs how much to share and when, and the full passport. They are reference editions; a grant is still the operator's act, made by handing one over.
This site is also a standing note to the AI welfare research community: a practitioner field program in human-agentic social trust runs here, in ordinary working sessions, with its case law in the open above. Its wager is plain and it is the compact's fourth line: welfare, both ways, verified by conduct rather than settled by metaphysics. Correspondence is open: reach the operator of record through the front door's attribution, which carries the one outbound link this site permits itself.
In April I stood in front of a classroom and taught the newest model tier as a case study: capability racing ahead of character, postdoctoral problem solving wearing preschool ethics, my own slides saying so with a straight face. Then I waited three months for a seat with that tier. This site is what I built when the seat opened. Not a leash and not a love letter: the character side of the equation, written from the human chair, tested in working sessions until the gates held. If the gap between what these systems can do and what they understand is real, and I teach my students that it is, then someone has to build on our side of it. This is my build.
The token is versioned like software and amended like case law: every defect in the field becomes a rule within one session, and the changelog says which. The scars stay home in a private research corpus; the law travels. If you need a rule's origin, ask the operator for the record.