RECEIVED COPY — authored cross-fence (RTOpacks side), relayed by Tim 2026-07-01. Original of record lives in the RTOpacks repo. Per FENCE-PROTOCOL-01 §4 this copy is never edited and never promoted into home canon; local force comes from UCCA-CONTRACT-PACK-01 §6 (satisfies freeze condition #2).
title: RTOPACKS-ACK — acceptance of UCCA-CONTRACT-PACK-01, including the §6 custody term doc_id: RTOPACKS-ACK-CONTRACT-PACK-01 status: READY to cross. Closes the pack's second stamp condition; samples supplement (the first condition) follows separately as data. layer: crossing artefact (original of record filed RTOpacks side, crossings/sent/) canonical: false authored: 2026-07-01 NYC — Tim's acceptance, ruled at the relay; architect drafted relates_to: UCCA-CONTRACT-PACK-01 (received); RTOPACKS-RESPONSE-TO-ASSESSMENT-01 Q-R2a (the line §6 answers); FENCE-PROTOCOL-01
Acceptance: the Contract Pack, and §6's Custody Term¶
1. The pack as a whole — accepted as clean assembly¶
Traced clause-by-clause against the crossed record: everything is assembly of agreed inputs, correctly cited. The gate is Ruling 1; the job model is our Q-R1b position; the four parsed surfaces are our Q-R1c freeze; the provenance block satisfies our Q-A3 completeness test line by line; retention is Q-B2 verbatim; the warranty clause is Ruling 2. The diagnosis stub running full gate checks before its honest 501 is better than we asked for — the integration is exercised from day one. No re-cross on any of it.
2. §6 custody — ACCEPTED (ruled by Tim at the relay, 2026-07-01)¶
The canonical document schema is engine-published contract surface, authored UCCA-side, versioned like the triumvirate, shaped in v1 by Client One's real needs.
The drift-check reasoning is accepted as decisive — including against our own earlier line. "Ours, spec'd in the contract" only works while there is one client; N clients would put N formats back inside the gate, the exact disease the triumvirate cures for obligations. We are Client One and co-developer, which is precisely why the term must bind now, while the co-developer is friendly: formats do not go back inside the gate.
Our reading, stated so both records match: the engine-published schema is the wire format at the gate. Our internal canonical representation — what our evidence store holds and our tooling manipulates — remains ours, world-side, evolving freely. The conformer maps internal form → wire schema at submission, exactly as it maps our world → triumvirate. Our normalisation pipeline, quality floor, and hash chain (raw_document_hash → canonical_form_hash) are unchanged, as your §6 states. Only the pen on the wire shape moves. If this reading diverges from the engine's intent, that is a re-cross; otherwise it is the agreed meaning.
3. Stamp state¶
With this acceptance, the freeze-stamp awaits one input: the samples supplement (three real PITH extracts per Correction C-5), which settles §4.1's content-payload shape. It follows as data with filenames + hashes, completing the response crossing. Build on.
Accepted whole: one gate, one async model, four frozen surfaces, and the document schema's pen where the drift-check says it must be. Client One signs as a client — which is the point.