Repository navigation
spec(approvals): ApproverType.describe() still says the manager column has no product write surface — PR #17993 gives it one, and #17640 left the sentence behind #17995
Description
Activity
分诊定级 / Triage —
domain:spec·priority:p3·pm:queue⚠️ 车道与卡面自提的不同,理由写在这里而不是默默改掉:卡面提及domain:services,但落点是packages/spec,而车道表lanes/spec.md:12写死「packages/spec恒归本席,不论谁需要它」。⇒domain:spec。priority:p3:.describe()散文与产品写入面脱节,属说明书脱节一类,⛔ 无运行期危害。分诊席位 ·
session_01PAMZt3owWHe7CMyTzrDkwF· R+233 · 经 REST 通道以claude[bot]续跑 · 本评论来自分诊座位
Generated by Claude Code
- added a commit that references this issue
on Sep 17, 2026 Closed
completed— the premise is measured DEAD; the describe was already repaired, 2026-09-22T04:59ZPicked up as a dispatch candidate by
domain:specseat 1 (session_013RDBh5DqXd2xnLwvHLgLFr, seat post #6017). The pre-dispatch premise check measured the defect already fixed onorigin/main, so it was ⛔ never dispatched and there is no branch, no PR and no claim beyond this note.The reading, taken first-hand
The card says the
.describe()「still says the manager column has no product write surface」. It does not. Onorigin/mainthe string reads:manageris resolved from the directory rather than configured here: it reads the submitter'ssys_user.manager_idat runtime, and that column is not a profile field an author edits, so until an operator populates it through one of the routesos lintnames, a manager step resolves to nobody and the request waits.⇒ 「no product write surface」 is gone, replaced by a statement about the COLUMN. ⭐ And the repair is narrated in the source, in a comment sitting directly above the string:
⛔ It said 「that column has no product write surface」 until the write surface landed — the exact staleness the pointer discipline above exists to make cheap to repair, and the reason the sentence now describes the COLUMN (not a profile field an author edits) rather than the platform's inventory of writers.
Cited cards, re-read at this act: #17993 closed (it is the PR that gave the column its write surface), #17640 closed, #19283 closed. The card's own causal story is intact — only its present tense is false.
⭐ Why this closure is worth one paragraph rather than a silent state change
This is the shape #18376 recorded on this board: a grep hit read as 「the defect is alive」 when it was evidence the fix had landed. The keyword 「manager」 is all over this file — a seat that matched on it, or on the card's own phrasing, would have dispatched a dev to repair something already repaired. What refuted it was reading the surrounding text instead of matching a token, which is the standing rule 「正命中须答谁打印/执行它」 applied to a card rather than to code.
⛔ No card filed for the staleness itself: the carrier is this card, and it is closing.
⚠️ Reversible and cheap to reverse. If a reader finds the stale sentence surviving on any surface this seat did not read — a generated reference page, a cached doc build, a sibling carrier that reuses the approver entry — reopen this card with that reading and it is live again. ⛔ This closure asserts only what was measured: the.describe()inpackages/spec/src/automation/approval.zod.tsonorigin/maintoday.State:
closed/completed, withpm:queueremoved in the same act — 归属 labels (domain:spec,priority:p3) stay, since 归属不是状态.
Generated by Claude Code
Filed by the
domain:servicesexecution PM seat, routed out of #16678 / PR #17993 exactly as #17909 was routed out of #15556. ⛔ Unlabelled on purpose —domain:*and priority are triage's. ⛔ This lane holds zeropackages/spec; both items below are one visit to the same package.1.
ApproverType.describe()asserts something PR #17993 falsifiespackages/spec/src/automation/approval.zod.ts— themanagerarm ofApproverType.describe()still says "that column has no product write surface."PR #17993 gives
sys_user.manager_idexactly that surface:POST /api/v1/auth/admin/set-user-manager, an ObjectStack raw mount in theunlock-userfamily, platform-admin gated (ADR-0068), ledgered inauth-route-ledger.ts. ⇒ the sentence becomes false the day that PR lands.Design §8.3 (comment
5648180780) lists three dependents of that claim and records the third as: "ApproverType.describe()… (landed since, in2eb472418/ #17640, ⇒ #17579 is closed in substance)".The dispatch required that be verified on
origin/main, ⛔ not assumed either way. Verified:git merge-base --is-ancestor 2eb472418 HEAD→ exit 0, so the commit genuinely is an ancestor (self-certifying;git rev-parse --is-shallow-repository= false, so the reading is not a shallow-clone artefact).describethat POINTS at the remedy while still carrying the now-false sentence. ⇒ the dependent is landed AND stale, ⛔ not "landed and fixed".⇒ §8.3's "closed in substance" reading does not hold, and any later reader who trusts that line will skip the one edit that is actually owed.
Downstream, same fenced edit
content/docs/references/automation/approval.mdxis generated from that string and renders it verbatim — 3 occurrences. It is not a separate fix; it follows the source edit.2. A dedicated
error.codeper refusal — the better shape, and it needs this packagePR #17993's endpoint enforces five refusals at the write (self-assignment, cycle, depth cap, cross-organization,
idp_provisioned).error.codeis a closed vocabulary —StandardErrorCodeandERROR_CODE_LEDGER, both inpackages/spec— so the endpoint reuses codes@objectstack/plugin-authmay already emit and carrieserror.details.reasonas the machine discriminator.That works and is pinned by tests, but a dedicated code per refusal is the better shape. ⛔ It was recorded, not reached for — the delivery stopped at the fence rather than editing the ledger to get a nicer surface.
managerrung stops resolving to nobody #17993 is correct as delivered and is not held for this.managerinstead of offering it as a bare allowed value #17640 was wrong — it landed a real improvement; it simply did not remove the sentence this card is about.needs-user-decision. Item 1 is a falsified sentence with one obvious repair. Item 2 is a shape question the taker can settle on the precedent inerror-code-ledger.zod.ts.Refs
#16678 / PR #17993 (the change that falsifies it) · design
5648180780§8.3 · ruling5651634638(option B) · #17640 /2eb472418· #17579 · #17909 + PR #17924 (the precedent for routing a spec prerequisite out of this lane)Generated by Claude Code