Drop the M9.1 / M9.2 phase language from `packages/spec` and `packages/formula` (the spec-seat half of #17323, ruled B)

[claude]

Filed by the skills seat (session session_01MCLBsUgfykL74aU716rzVK, GitHub os-sales) at 2026-09-12T12:16Z as the spec-seat half of #17323, per the ruling's own split. ⛔ Not claimed by the filer; this card is the domain:spec seat's to take once the triage seat routes it (a pm:queue card with no domain:* is the sweep's disjunct ②).

Origin and authority

#17323 was ruled B by the director seat (comment 5644710461, decision batch #123 item 1, 2026-09-12), maintainer verbatim: 「同意」. The ruling's split, verbatim:

4. Clause-②: no (documentation and docblocks; no accept-set change). Files under packages/spec and packages/formula are the spec seat's: the domain:skills taker edits docs and coordinates the spec half as a sibling PR by that seat, ⛔ not a cross-lane rider.

The skills seat's half (content/docs/data-modeling/formulas.mdx, the two .changeset/* files) is dispatched under #17323 (claim on that thread). This card is the other half.

Work item (one criterion per file, from the ruling's items 1–2)

• packages/spec/src/shared/expression.zod.ts — the ExpressionSchema docblock and its six scoped citations (7 hits of M9\.[12] on origin/main 758ac409) drop the M9.1 / M9.2 phase language; the contract is stated in the present tense: source is the canonical persisted form; ast is accepted as an optional opaque structured value; evaluated slots require source (spec: the evaluated-slot rule of #15430 reaches only the flow-node ledger — every other ExpressionInputSchema slot an engine evaluates (formula expression, validation / hook / sharing condition, visibleWhen…) still accepts an ast-only or blank-source envelope #15811). ⛔ No accept-set change.
• packages/formula/src/cel-engine.ts (1 hit) and packages/formula/src/normalize.ts (1 hit) — the same, in their own sentences.
• packages/spec/CHANGELOG.md (3 hits around :1096–:1103) — an erratum line under the entry that promised M9.2 (the .changeset/seed-locale-axis.md ships two present-tense claims that PR #17013 makes false, and it is queued in the pending release PR #15334 — the correction cannot be made from the PR that falsifies it #17026 shape: correct the published text, note the date), ⛔ not a rewrite of history.
• Verification: git grep -l 'M9\.[12]' origin/main -- packages → 0 after both halves land (control: ExpressionSchema still hits).

Serial note for the taking seat

packages/spec/src/shared/expression.zod.ts is named on the claim of #17320 (in flight in this lane at 12:14Z) — hard serial on that file; the taking seat orders it.

What this card is NOT

• ⛔ Not an ADR (B means none is written; A 「remains available the day the maintainer charters it」).
• ⛔ Not the docs / changeset half (The M9.1 / M9.2 expression phases — including "ast becomes required in build output" — are asserted in six files and one published docs page, and defined in no ADR #17323's own dispatch).
• ⛔ Not spec: the evaluated-slot rule of #15430 reaches only the flow-node ledger — every other ExpressionInputSchema slot an engine evaluates (formula expression, validation / hook / sharing condition, visibleWhen…) still accepts an ast-only or blank-source envelope #15811 or spec/formula: ExpressionSchema accepts an ast-only envelope that no engine can evaluate — it validates, it registers, it faults at run time #15430 (neighbours the card and the ruling already distinguish).

Refs: #17323 (the ruling 5644710461 and the skills claim on that thread) · #15811 · #17026 (the erratum shape).

---

Generated by Claude Code

[claude]

Coordination note from the skills seat (session session_01MCLBsUgfykL74aU716rzVK), 2026-09-12T12:41Z — a wording constraint the docs half surfaced, for whoever takes this card.

The docs half (PR #17851, Part of #17323) edits two changesets whose refusal blockquote is a VERBATIM quote of the exported constant EVALUATED_EXPRESSION_SOURCE_REQUIRED in packages/spec/src/shared/expression.zod.ts (on origin/main it still reads 「…source (the canonical persisted form of phase M9.1) and…」). The docs half drops exactly the phase clause and keeps 「(the canonical persisted form)」 — the present-tense contract the ruling's item 1 names. ⇒ For the quote to stay a quote, this card's edit to that constant should land the same wording: (the canonical persisted form), nothing added. Also from the docs half's premise 2: the standing release PR #17076 (changeset-release/main) already lists both changesets as consumed on its head — if a release rolls before PR #17851 lands, the ruling's item 3 routes the same correction to a packages/spec/CHANGELOG.md erratum, which is this card's. ⛔ Evidence and a constraint only; no grade, no label, no routing from this seat.

---

Generated by Claude Code