Skip to content

[finding] driver-memory keeps a THIRD refusal wording for the same dateRange condition — the tree is at two wordings, not one, after #18232 #18277

Description

@claude

Reported by the domain:services dev that delivered #18275 (card #18232) as an out-of-scope finding, and filed here by the seat.

⚠️ Filed by the seat because the dev's dedupe control FAILED, and that is stated rather than hidden. Its one targeted semantic search did not return #18232 itself — the known-must-hit control for this exact subject — so the empty result was not a reading, and it refused to file on it. ⭐ That refusal is correct: an empty result from an instrument that cannot be shown to hit a known positive is worth nothing. ⛔ This card therefore carries NO dedupe claim; establishing whether a duplicate exists is part of triage's work here, not a step this seat is pretending was done.

The finding

packages/drivers/driver-memory/src/memory-analytics.ts has its own local explicitDateRangeWindow that overwrites err.message with a [driver-memory] sentence — a third wording for the same condition, under the same ANALYTICS_DATE_RANGE_UNRECOGNIZED envelope.

The rule it violates is quoted in packages/spec/src/data/analytics.zod.ts's own prose, the #5240 convention:

the refusal is shared by the schema door (this file) and, through the ANALYTICS_DATE_RANGE_UNRECOGNIZED envelope, by the runtime door and the drivers (#16322), so one condition keeps one wording

⚠️ Note "and the drivers" — the convention names this package explicitly.

Why it is filed NOW and what it means for #18232

PR #18275 removes the service-analytics face's second wording. ⇒ after it lands the tree is at TWO wordings for this condition, ⛔ not one. Card #18232 asked for the convention to hold, and it will not hold repo-wide until this one is closed too.

⛔ It is not a defect in #18275, and ⛔ it was correctly not swept in: packages/drivers/** is outside that card's declared file surface, and a one-off sweep from a card that does not own the package is how a shared convention gets moved sideways instead of forward.

Class (b): the same declared invariant, violated the same way, in a different package.

Successor

Whoever holds domain:engine / driver-memory. This is the exact sibling of the card just delivered, so the shape of the fix is already demonstrated in PR #18275.


Dedupe words: driver-memory memory-analytics explicitDateRangeWindow third wording #5240

⛔ Ungraded on purpose, and ⛔ no domain: label — the surface is packages/drivers/**, which is not this seat's to route. Grading and routing are triage's.


Generated by Claude Code

Activity

  1. objectstack-fleet commented on Sep 23, 2026

    @objectstack-fleet
    Contributor

    关 not_planned —— 维护者逐张复核 on-hold 卡时同意关闭;剩下的第二种措辞在冻结中的内存驱动里

    分诊席(session_01Tw7jnJinGHvoGSi8aFkhPJ),2026-09-23T11:19Z。维护者 2026-09-23 在分诊会话里逐张复核 pm:on-hold 卡,对第九组的回复原文:「第九组 同意」。本卡在那一组里的建议是关闭。

    现状

    为什么关

    给以后的人

    若内存驱动的这一段因为别的原因被修改(或冻结解除),顺手照 PR #18275 的样子去掉本地覆盖,让它沿用共享的那一种措辞。

    关闭理由:not_planned,同时摘掉 pm:on-hold。


    Generated by Claude Code

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions