[finding] The "maintainer 2026-08-27: no alias, no deprecation window" ruling cited by #16483 and PR #16665 cannot be located, and its date contradicts the "same week" prose that cites it

[os-sales]

Raised by the at-tier (fable) contract review of PR #16665 (finding F1), dispatched by the domain:cli execution PM seat (#6024). Filed by that seat. Unassigned, bare, for triage.

⛔ This card does not decide the ruling, and nothing here questions the retirement itself. The contract review's verdict on #16665 was PASS-WITH-FINDINGS and F1 was explicitly not blocking — the substance was verified against the shipped os g agent precedent. What cannot be verified is the citation.

The claim

#16483 and PR #16665 both attribute the no-alias / no-deprecation-window decision to a maintainer ruling of 2026-08-27. That attribution has propagated into six places:

• the card body (cli: retire os create example — it is a weaker os init plus a README; docs point at os init (item 2 of #15531) #16483)
• PR feat(cli)!: retire os create example; the refusal names os init (#16483) #16665's body
• packages/cli/src/commands/create.ts:13-14
• the header of packages/cli/test/create-example-retired.e2e.test.ts
• .changeset/os-create-example-retired.md
• the retirement callout in content/docs/deployment/cli.mdx

⚠️ It could not be found — searched, not assumed

The reviewer searched:

• all six comments on [finding] Measure whether os create and os init should converge — the shared surface is now four exports and one file map away #15531 — the batch Complete documentation structure optimization with guides, standards, and templates #66 ruling (director seat, 2026-09-07, option B, comment 5564757555) says verbatim only "Retire os create example (a weaker os init plus a README) — cli: retire os create example — it is a weaker os init plus a README; docs point at os init (item 2 of #15531) #16483". No alias or window language.
• all 200 comments on ledger [PM seat] director(项目总监) — 🟢 os-zhuang（第 15 场，契约复审 + 落地）· ⏹️ hotlong（第 17 场已收班）· 第 16 场已收班 #12708 — the batch Complete documentation structure optimization with guides, standards, and templates #66 row matches [finding] Measure whether os create and os init should converge — the shared surface is now four exports and one file map away #15531 and carries no alias/window language. ⭐ The ledger's earliest comment is 2026-08-30, so it cannot carry an 2026-08-27 ruling at all.
• the head tree (AGENTS.md, docs/, .claude/, scripts/pm, .changeset/, packages/cli/src, content/docs) for "deprecation window" and "2026-08-27". Every 2026-08-27 ruling found is about something else: [decision] Does the v5.0 project → environment rename extend to the API surface — client.projects.* and the res.project / res.projects response fields? #12473, [Decision] Is ADR-0112's declaredCode channel in scope for 5xx sanitisation at all — and the answer must be applied to all three doors at once #12509, The two ⛔ notes forbidding a single-sourced ORGANIZATIONS_RUNTIME_PKG name a constraint that no longer exists — decide whether to consume the resolver hop or restate why not #12579, CryptoContext.namespace / .key document themselves as settings-only, one layer below sys_secret's three producers #12599, Surface licensed max_nodes oversell to operators — visibility without the slot-claim mechanism (ruled C on #8501) #12667, approvals: escalation.enabled has zero readers — the SLA sweep gates only on timeoutHours, so enabled: false still escalates #12278, PluginConfigValidator can never run: PluginLoader calls its own validatePluginConfig(metadata) with no config, and a plugin factory closes over its config so the kernel never receives it #11982, objectui#6262.

⭐ Two internal tensions — one statement is wrong whichever way this resolves

1. 2026-08-27 predates the measurement that first identified the duplicate. The os create example / os init convergence was measured by PR docs(audits): measure whether os create and os init should converge #15797 on 2026-09-05. A ruling to retire the command could not precede the measurement that found it was a duplicate.
2. create.ts:13-14 and the e2e header say the maintainer "ruled the same week" as [finding] Measure whether os create and os init should converge — the shared surface is now four exports and one file map away #15531, which is dated 2026-09-07. 2026-08-27 is not that week.

⇒ The date and the "same week" prose cannot both be right.

What is being asked

Confirm the ruling and its date. Then the "same week" prose in create.ts and create-example-retired.e2e.test.ts should be corrected to the real date.

⚠️ What has already been done, so this is not re-done

PR #16665's dev has been instructed by the PM seat to make the citation assert only what is verifiable — cite the #15531 batch #66 ruling (verified verbatim by the review) and record the alias/no-window provenance as recorded on card #16483, pending maintainer confirmation, dropping the unsupported date and the "same week" phrasing.

⛔ The runtime message was deliberately left alone: "There is no alias and no deprecation window" describes what the code does and is true regardless of who ruled it. Only the provenance claims were narrowed.

⇒ So the PR does not ship the unverifiable date. This card exists so the underlying question survives #16483 closing on merge, and so the real date can be restored if one exists.

Why it is worth a card rather than a shrug

This repo's whole os create example retirement is about a statement that outlived its truth. A ruling citation that six artifacts repeat and no one can locate is the same shape, one level up — and it is load-bearing, because "no alias, no deprecation window" is the part a consumer is told is not negotiable.

Refs: PR #16665 · #16483 · #15531 (batch #66, comment 5564757555) · #12708 (ledger) · PR #15797 (the 2026-09-05 measurement).