decision: should the losslessly mappable subset of the record-form filter get a real D2 conversion, or is "stored rows keep loading, re-saving is refused" the final answer?

Filed by the domain:spec execution seat, session_01MkQhmuuJAVDjmeWNixwDDH, 2026-09-10T06:40Z, out of the at-ACCEPT residue of PR #17257 (#15442 / #15449). ⛔ Unclaimed. No domain:* label and no pm:* state: ⛔ both are the triage seat's to produce. This is a decision card by shape — the four-facet block below is filled in at filing time as the protocol requires; ⛔ the seat does not adjudicate it.

Measured on origin/main 501959b72.

One sentence

A customer who saved a page last month with a filter written the old way will find it still renders exactly as before — right up until someone opens that page in the builder and hits Save, at which point the platform refuses the save and hands back a message that does not say how to fix it, and there is no tool that will fix it for them.

Background

Seven filter doors converged on the rule array [{ field, operator, value }, ...] (objectui#6206, maintainer batch adjudication 2026-08-25, verbatim 「同意」, Option B; family-wide convergence, decision batch #55, 2026-09-06, verbatim 「同意」, option A). The shipped disposition is deliberate and is written into the element-data-source-and-object-block-filter-rule-array entry in its own words:

⚠️ Metadata AT REST is deliberately NOT rewritten, and this disposition adds no D2 conversion — a SemanticMigration converts nothing by its own type, and os migrate meta --stored (the pass over a deployment's sys_metadata rows) replays D2 conversions only, so it has nothing to rewrite for this shape. The read path does not re-validate stored rows (applyConversionsToStoredItem replays the full chain without validating, by its own contract), so a stored page or block carrying the record form keeps loading unchanged and is still rendered by objectui at the pinned .objectui-sha; what changes is that RE-SAVING it is refused at the filter door, on its next save and not before.

So today: read = fine, re-save = refused, no conversion, no rewriter.

The question the convergence left open

The three entries' own replacement text already spells out a total, mechanical mapping for the common shapes:

flat record — { status: 'active' } → [{ field: 'status', operator: 'equals', value: 'active' }]
operator object — { amount: { $gt: 100 } } → [{ field: 'amount', operator: 'greater_than', value: 100 }]
several keys → several rules (they AND)
ObjectQL AST tuple array — [['owner_id', '=', '{current_user_id}']] → [{ field: 'owner_id', operator: 'equals', value: '{current_user_id}' }]
legacy operator shorthands (eq, ne, gt, notIn, …) accepted and normalized on parse

That mapping is losslessly executable by a machine for exactly that subset. What it is not total over is $and / $or / $not — the record's combinator keys — because the rule array is a flat AND-list with no nesting.

⚠️ The one measured fact that decides how much is at stake: the record's $and / $or / $not keys "were misread by every gate block anyway (objectui#6948)" — that is the shipped entry's own words, and objectui#6948 is closed. So the nesting the rule array cannot express is nesting the consumer never honoured. That makes the un-mappable remainder small, but it is exactly the part where a mechanical conversion would have to either refuse or change behaviour.

Governing text

ADR-0087 §D2 (a real conversion, replayed by os migrate meta --stored) vs §D3 (SemanticMigration — prose, converts nothing; packages/spec/src/migrations/types.ts).
ADR-0087 at its own not-required (no-migration-prescription) clause refuses that disposition when the entry body carries a prescription. These three entries carry a full prescription, so D3 is the floor, not not-required.
packages/spec/src/conversions/stored.ts — applyConversionsToStoredItem replays the chain without validating, which is why stored rows keep loading.

Restart-touch: packages/spec/src/migrations/entries/semantic/18.element-data-source-and-object-block-filter-rule-array.ts

Re-check before acting: git log --oneline -20 origin/main -- packages/spec/src/migrations/entries/semantic/ and git grep -n "'element-data-source-and-object-block-filter-rule-array'" origin/main -- packages/spec/src/migrations/registry.ts — if a D2 has been added since, this card is already answered.

Options and what each actually costs

	what gets built	what the customer sees
A — leave it (the shipped disposition)	nothing	Old pages keep rendering. The first person to re-save one gets a refusal. Whoever that is has to hand-rewrite the filter, with the current bare `invalid_type` as their only guidance (that message gap is a separate card). Nothing tells them how many such pages a tenant has.
B — partial D2: convert the mappable subset, leave combinators alone	a real D2 conversion covering flat records, `{k: {$op}}` objects and single-level AST tuples; a combinator-carrying record is passed through untouched	`os migrate meta --stored` cleans up the great majority of stored rows in one operator-run pass; those pages re-save cleanly afterwards. A tenant with a `$and`/`$or` filter is left exactly where option A leaves everyone — but that is a much smaller population, and (measured) that nesting was never honoured downstream anyway. Cost: a conversion that is deliberately partial is a shape this ledger has not shipped before, so "did it convert my row or not" becomes a question an operator can ask and today cannot answer.
C — full D2, converting combinators by flattening them	as B, plus `$and` flattened into the AND-list and `$or` / `$not` handled by some chosen rule	Every row converts; nothing is left behind. But `$or` has no representation in a flat AND-list, so this option necessarily changes what some pages select — silently, during an operator-run migration, on data. That is the failure mode this platform's rules exist to prevent.

Business translation

A = "we changed the lock and did not issue new keys; whoever tries the door next finds out."
B = "we re-key every door we can re-key mechanically, and put a note on the handful we cannot."
C = "we re-key every door, and accept that a few of them now open onto a different room."

os-decision-facets

① 项目长远合理性:B 缩小特例 —— 它把「旧写法怎么办」从每个客户各自的意外,收敛成一次可运维的批处理加一份点名清单;A 把这个特例永久留在客户侧,每张老页面都是一次将来的意外;C 表面上最干净,但它引入的是「迁移会悄悄改变筛选结果」这条永久债务,比它清掉的特例贵得多。
② 实际业务拉动:今天撞上的人 = 任何一个打开旧页面按保存的人。⚠️ 拉动的真实大小本席测不到 —— 那是各部署 sys_metadata 里的存量,不在仓里;仓内作者面已在同一个改动里全部改写(实测:绑定处 4 个 spec fixture、block 门 5 个 showcase 作者、3 个 lint fixture,仓外为零)。⇒ 若维护者判断线上存量接近零,零拉动的默认动作是 A。
③ 防 AI 犯错:A 的失败是响亮的(保存被拒),这是它最强的一面;它的弱点是拒绝话术不说怎么改(另卡)。B 同样响亮,且把响亮的次数降到最低。⚠️ C 的失败是静默的 —— 迁移跑完不报错,页面照常打开,只是选出来的行变了。⇒ 单凭这一轴,C 应当出局。
④ 创业阶段不扩散:A 不新增任何长期义务;B 新增一个必须长期正确的转换器,但它转换的是一个已经冻结的旧形状(接受集已经关了),所以这份义务是有终点的、可删除的;C 新增的是一份永远说不清的义务。⇒ 不扩散这一轴上 A > B ≫ C。

推荐:B(部分 D2),回退项 A。理由:它是唯一一个既降低客户侧意外、又不把静默行为改变引进迁移路径的选项;而它新增的义务有明确终点。⚠️ 置信缺口 —— 本分析看不见的东西:线上部署 sys_metadata 里究竟有多少行携带旧记录形筛选,以及其中多少携带 $and/$or/$not。这个数一旦接近零,推荐立刻翻成 A;它只能由维护者或一次面向部署的读数回答,⛔ 仓内测不出来。

维护者速读

老页面里用旧写法存下来的筛选条件,现在读没问题、存会被拒,而且平台不提供任何自动修复。绝大部分旧写法可以被机器一对一改写成新写法;只有带「或者 / 并且 / 非」这种嵌套组合的那一小撮不行 —— 而实测表明,那种嵌套前端从来就没真正执行过。

A:什么都不做,谁碰到谁自己改。
B:写一个只改能改的那部分的转换器,运维跑一次批处理清掉绝大多数,剩下那一小撮点名列出来。(推荐)
C:全都自动改,包括嵌套那一撮 —— 代价是有些页面迁移后筛出来的数据会悄悄变。

⚠️ 决定这件事划不划算的唯一数字,是线上到底有多少张老页面带旧写法的筛选 —— 这个数只有你看得到。若接近零,选 A。

请回一个字母:A / B / C。

The card's confidence gap says the recommendation flips to A if the on-deployment count of old-form filters is near zero, and that the number "⛔ 仓内测不出来". That is right — and it is the same wall this seat hit on objectui#8365 (stored kanban views) and objectui#8894 (multi-measure dashboards) today. ⭐ Three decision cards in one round turn on a number that only a production read can answer. ⚠️ If the maintainer has a way to run one such read, answering all three at once is far cheaper than three separate rulings.

⛔ This seat does not adjudicate, does not narrow the options, and does not touch the recommendation.

_{分诊席位 · session_017VGfRocA8VjczSe84fgjY3 · R+166 · 2026-09-10T14:36Z · 本评论来自分诊座位}

Generated by Claude Code

Ruling recorded — B: a partial D2 conversion for the losslessly mappable record-form filters; combinator-carrying rows pass through untouched and are named (director seat, decision batch #121 item 4, 2026-09-12)

Maintainer, verbatim (live PM chat, 2026-09-12T04:5xZ), to decision batch #121 presented as 1C+D·2A·3A·4B·5乙: 「同意」.

Derived first from the long-term axis: a migration converts what it can convert losslessly and names what it cannot — that is ADR-0087 D3's own model (conversion where lossless, structured TODO otherwise), so B is not a new shape for this ledger, it is the standard one. A leaves the special case on every customer's next save; C converts combinators by flattening and silently changes what some pages select — excluded on the fault axis alone.

What is ruled

A D2 conversion in the element-data-source-and-object-block-filter-rule-array family covering: flat record {k: v} → [{field, operator: 'equals', value}]; operator object {k: {$op: v}} → the mapped operator; several keys → several AND rules; single-level AST tuples where the seat's mapping is total.
A record carrying $and / $or / $not is passed through unchanged and reported as a structured TODO naming the page/block and the combinator — os migrate meta --stored prints the list; the operator can answer "did it convert my row" from that output.
The refusal message at the filter door gains the "how to fix" sentence (the separate message-gap card is folded in if it is still open; otherwise cross-referenced).
Clause-②: no (no accept-set change; a conversion is added to the ledger).

State

needs-user-decision → pm:queue; domain:spec / priority:p2 kept.

Generated by Claude Code

feat(pm): mechanise the prior-ruling search for decision cards (objec…

Claim: PM loop round 1 · 2026-09-27T03:45Z
Session: session_01Rjy9MeetSfq34PKn81CRiN
Account: os-zhuang
Branch: claude/issue-17321-record-filter-d2-conversion
Worktree: objectstack-issue-17321
Domain: domain:spec
Seat: domain:spec#1
File surface: packages/spec/src/conversions/ (the D2 registry, the new conversion and its tests); packages/spec/src/ui/filter-rule-array.ts (only if the mapping must share the doors' one operator table); the family's ADR-0087 semantic entries whose prose this makes false (packages/spec/src/migrations/entries/semantic/18.*filter-rule-array*.ts) + registry; the generated artefacts and reference pages that follow, regenerated and never hand-edited; .changeset/. A structured-TODO change outside packages/spec is stopped and reported, ⛔ not widened into. (stop on breach; explain in the report)
Container & model: M, mode:subagent, model: default judgment tier (dispatch-gates.mjs --tier at 9401b842: 「no path-derived mandate」, the default slot taken; the diff hits the enqueue gate's path limb, so it is reviewed at CONTRACT_REVIEW_TIER before enqueue). Verify-lock arrival depth 1 at the gate read.
Clause-②: no
Thread-read: 5644018752
Serial constraints cleared: Ruling-ref: 5644018752 (B, batch #121 item 4), retrieved this act. Ruling item 3's message-gap card is #17320, closed completed 2026-09-12; it is cross-referenced, ⛔ not redone. No open PR (of all open PRs, file lists read this act) touches packages/spec/src/conversions/, ui/filter-rule-array.ts, ui/page.zod.ts or ui/component.zod.ts. migrations/registry.ts: ordinary concurrency, regenerated. No prior claim on this card. Clause-② no, as ruled: no accept-set change; a conversion is added to the ledger.

os-dev-report
{
"issue": 17321,
"status": "done",
"branch": "claude/issue-17321-record-filter-d2-conversion",
"pr": "#20175",
"session": "session_01Rjy9MeetSfq34PKn81CRiN — mode:subagent, the PM session; identity = the branch named by Claim 5852390101 (verified newest Claim names this branch before any code)",
"premise_still_valid": true,
"summary": "Ruling items 1, 3 and 4 are delivered, and so is the pass-through half of item 2, in packages/spec only. Draft PR #20175 is at head b1c9143. Item 1: new ADR-0087 D2 conversion page-component-filter-record-to-rule-array (toMajor 18, retiredFromLoadPath: true, in step18.conversionIds). It rewrites a flat record, an operator object (several operators on one field become several rules), several keys and a single-level AST (one comparison, or a flat list of them) into the ViewFilterRule array wherever the mapping is lossless. It reaches 12 doors, measured door by door and pinned against the schema: dataSource.filter on any page component; properties.filter on object-grid, object-metric, object-kanban, object-calendar, object-map, object-gantt, object-tree, object-timeline, element:number and element:record_picker; and object-grid properties.defaultFilters. The operator comes from the existing tables and nothing else: FILTER_OPERATORS plus normalizeFilterOperator, and AST infix spellings lowered through parseFilterAST. Every produced rule must parse at the door (ViewFilterRuleSchema). Losslessness is measured per operator over FILTER_OPERATORS and VALID_AST_OPERATORS against parseFilterAST. Item 2: $and, $or and $not, AST and/or groups, and every shape without a lossless spelling are left byte-identical. That covers null values (measured at the objectui pin f8a9d0fb: convertFiltersToAST skips a null key), $null, $exists, like and ilike, and comparands the door refuses. Conversion is all-or-nothing per filter, and nothing is ever flattened. The printed TODO list half of item 2 is STOPPED as the dispatch ordered, because the existing machinery does not report pass-through rows. Sites: (1) spec conversions/types.ts, apply.ts and index.ts have no TODO channel. (2) packages/runtime/src/dispatcher-error-vocabulary.ts: check:dispatcher-error-vocabulary would go red on a new code stamp in apply.ts until a foreign-vocabulary row is added, so the spec half cannot land alone. (3) packages/metadata-protocol/src/protocol.ts: convertStoredItemDetailed forwards only onNotice, and migrateStoredMetadata sets changed = notices.length > 0, so a TODO-only row is recorded canonical. (4) packages/metadata-protocol/src/stored-migration.ts: StoredMigrationRow and formatStoredMigrationReport have no TODO field or section. The cli runStored prints that report, so it needs no change once the report carries the list. Item 3: #17320 is closed (completed) and is cross-referenced. All 12 doors already give the how-to-fix sentence. One gap was fixed: for a record carrying a combinator, the worked rewrite rendered the combinator as a field. Now it names the combinator and says why no rule spells it, and the text for every other value is byte-identical to base (11/11 inputs). The ADR-0087 prose made false by this PR is corrected in element-data-source-and-object-block-filter-rule-array and object-grid-default-filters-rule-array, with the registry regenerated; the new prose says in so many words that os migrate meta --stored does not list the rows left as stored yet. The PR body opens with Part of #17321, not the dispatch's Fixes, because item 2's list half remains (see deviations).",
"tests": "All at head b1c9143 (after merging origin/main 84880f9). [build] pnpm --filter @objectstack/spec build: VERDICT command-exit 0, dts 36/36. [generated] spec check:generated: all 15 up to date (spec-changes and upgrade-guide unchanged because protocol 18 is unreleased; PROTOCOL_VERSION is 17.0.0). [typecheck] pnpm --filter @objectstack/spec typecheck: VERDICT command-exit 0; check:test-typecheck OK with 142 pinned signatures and no new debt. [test] vitest run --project local --maxWorkers=2 over packages/spec: Test Files 541 passed (541), Tests 15863 passed | 2 todo. New file conversions/page-component-filter-record-to-rule-array.test.ts: 49 passed (sections 0-7: exact rule arrays per ruled shape and door kind; combinator and declined rows byte-identical with no notice; identity and idempotence; losslessness over FILTER_OPERATORS, declined = [$exists, $null], and over VALID_AST_OPERATORS, declined = [ilike, like]; the doors accept the output and refuse the source (control); reach equals the schema-derived door set for filter and defaultFilters; retired from normalizeStackInput, replayed by applyConversionsToStoredItem and by applyMetaMigrations(…, 17, 18)). filter-rule-array-guidance.test.ts: +2 combinator pins (code invalid_type + path filter + text). [ablation] node scripts/ablation-replace.mjs wrap mode on committed head, under the verify lock: anchor "return rules.every((rule) => ViewFilterRuleSchema.safeParse(rule).success) ? rules : undefined;" replaced by "return rules;", anchor 1->0, blob 17b8a6c1->188a52b8; expected direction red; observed 3 failed | 46 passed (field-reference comparand, in with a scalar, empty icontains); restore blob == HEAD 17b8a6c1 and git diff HEAD empty. The fourth door-refused row stayed green because parseFilterAST declines it first. [message byte-identity] base 49144fc filter-rule-array.ts loaded beside head over 13 inputs: identical=11 changed=2 (only the combinator inputs); the temp file was removed and the tree is clean. [consumers] @objectstack/metadata-core artifact-forward-conversion.test.ts 25/25; @objectstack/lint validate-component-props + validate-preset-comparands + authoring-rule-input-tier 107/107 (built via turbo run build for formula, lint and metadata-core, VERDICT 0). No test outside packages/spec seeds a page row with a record-form filter at these doors, and none pins the conversion list (swept all consumer tests of migrateStoredMetadata and applyConversionsToStoredItem). [smoke] built CJS and ESM dist: applyConversionsToStoredItem converts, and the CJS dist/ui combinator refusal text is quoted in the PR body. [lint, narrowed] eslint --no-inline-config --format json over 7 changed TS files: 7 files, 0 errors, 0 warnings; population = eslint.config.mjs packages//*.ts globs; no parserOptions.project and no typed rules, so there is no cross-file verdict movement. [cli integration tier] not touched; declared to CI.",
"gates": "dispatch-gates.mjs --commands --repo objectstack-ai/objectstack re-derived at b1c9143: 87 families. Each run was recorded as command :: exit code, captured before any pipe. 84 exit 0; 3 exit 3 NOT MEASURED, reason prerequisite not met: check:dual-build-cjs-loads and check:type-check-debt need a full pnpm build, and check:lean-entry-closure needs the objectql dist. check:doc-formula-expressions first exited 3 and exited 0 after building formula and lint. --ran reconciliation: 87 derived, 84 run, 3 NOT-MEASURED (derived from recorded exit 3), 0 UNRUN. Roster gates whose roster sits under my paths, all exit 0: check:filter-alias-parity, check:error-code-casing, check:authz-resolver, spec check:meta-url-spelling, spec check:error-code-provenance, check:nul-bytes. CI on PR #20175 at b1c9143 at report time: 32 check runs, 15 completed with 0 failures, 17 in_progress.",
"line_budget": "n/a: no skills/ file or line-ratcheted governed ledger is touched (governed surfaces untouched; diff is packages/spec/src/** + .changeset only).",
"files_changed": [
".changeset/17321-record-filter-d2-conversion.md (new, @objectstack/spec minor)",
"packages/spec/src/conversions/registry.ts (+ conversion, mapper, door sets, imports)",
"packages/spec/src/conversions/page-component-filter-record-to-rule-array.test.ts (new)",
"packages/spec/src/ui/filter-rule-array.ts (isRecordForm exported, not re-exported publicly; combinator branch of the refusal)",
"packages/spec/src/ui/filter-rule-array-guidance.test.ts (+2 pins)",
"packages/spec/src/migrations/entries/semantic/18.element-data-source-and-object-block-filter-rule-array.ts (at-rest prose)",
"packages/spec/src/migrations/entries/semantic/18.object-grid-default-filters-rule-array.ts (at-rest prose)",
"packages/spec/src/migrations/registry.ts (step18 conversionIds + rationale sentence; generated regions regenerated)"
],
"deviations": [
"PR body line 1 is Part of #17321, not the dispatch's Fixes #17321. Ruling item 2's printed-TODO half is stopped per this dispatch's own order, and os-dev.md says to use Part of when merging should not close the card. The seat can switch it to Fixes if it files the remainder as its own card (see open_questions).",
"Ruling item 3: the dispatch said add a sentence only where one is missing. I read the combinator case as missing, because the existing sentence prescribed a field literally named $or. The refusal changed for combinator records only; the text for every other value is byte-identical to base (measured 11/11).",
"The family was read as the 12 doors that ruleArrayFilterError serves, which is wider than the 5 doors of the named entry: element:number, element:record_picker, object-grid.defaultFilters and object-map/-gantt/-tree/-timeline are included because they share the same prescription helper, the same record form and the same conversion table. Measured and pinned in section 6.",
"I merged origin/main (84880f9) into the branch before final verification, because main had moved packages/spec migrations registry and report.zod. Main has moved again since (e2c4e12), touching none of the 8 changed files, so it was not merged a second time; the CI merge ref covers it.",
"I killed one of my own queued verify-lock waiters (pid 4198, recorded) to merge main first; the slot kept its place. No other process was touched."
],
"mcp_calls": "0 — no MCP GitHub tools called.",
"api_writes": "3 relay writes, each ONE POST /repos/objectstack-ai/objectstack/dispatches executed by the fleet-write workflow as objectstack-fleet[bot]: (1) pr_create → POST /repos/objectstack-ai/objectstack/pulls (draft, #20175, run 36296972010); (2) label-write --assign os-zhuang → POST /repos//issues/20175/assignees (run 36297019040, read back as matching); (3) post-stamped → POST /repos//issues/17321/comments (this os-dev-report). Plus git push of the branch (the probe push and 5 pushes; not REST). Reads: GET REST for the card, its comments, the PR and check-runs.",
"open_questions": [
{
"question": "Where should the stopped half of ruling item 2 live? That half is the structured TODO list that os migrate meta --stored prints, and it needs a spec TODO sink with a new code, a foreign-vocabulary row in packages/runtime dispatcher-error-vocabulary.ts, and the metadata-protocol stored pass and report.",
"options": [
"A: keep #17321 open as its carrier (PR #20175 stays Part of #17321), and dispatch the remainder against #17321 in a later round to a lane that owns metadata-protocol and runtime.",
"B: the seat files one new card for the TODO channel, with the 4 sites from this report and PR #20175 as its prerequisite, as a sub-issue of #17321, and switches PR #20175 line 1 to Fixes #17321."
],
"recommendation": "B. On the long-term axis, the remainder is a cross-package mechanism (a generic partial-conversion TODO channel) that should be decided and reviewed on its own, not as residue of a spec card. On the AI-error axis, one card per mechanism keeps what is left explicit. On the startup-focus axis, B adds no scope, only reshapes it. On the real-business axis, the pull is only the combinator rows, which today still load and are refused on re-save with a correct prescription, so the list is not urgent."
}
],
"out_of_scope_findings": [
"carrier: 承接者:无 · noted, not filed. The pending changeset .changeset/19514-view-filter-rule-scalar-arm-and-icontains-comparand.md says of object-grid.defaultFilters that stored metadata keeps loading and that re-saving it is refused. After #20175 that is still true for the forms left as stored, but not for the converted record and AST forms. #20175's own changeset states the new behaviour in the same release. This is not in the PR Acceptance notes and should be added by the seat if wanted.",
"carrier: 承接者:无 · noted, not filed. A non-combinator top-level $ key in a record (for example $text) is still rendered by the door's worked rewrite as field: '$text', and ViewFilterRuleSchema.field is a bare z.string(), so the door would accept that rule. It is unmeasured at a public door (reach not established), so it is not classed (c). Dedupe words: ViewFilterRule field dollar key, rule-array refusal worked rewrite, record-form $-prefixed key.",
"carrier: 承接者:无 · noted in PR Acceptance notes. $null and $exists operator objects are not converted, because they depend on the value and are not in the one table, and no report lists them until the stopped TODO half lands."
]
}

Generated by Claude Code

Review: ACCEPT (part) · PR #20175 at head `b1c9143f9ede26bfb8cea990f1b43cb86ac2ba38` · 2026-09-27T05:27Z

domain:spec seat 1 (session_01Rjy9MeetSfq34PKn81CRiN), reviewer of record, on claim 5852390101. Checked against GitHub, ⛔ not against the report (5852968339).

Checklist

Shape: draft, base main, first line Part of #17321. A full-body scan finds no closing keyword against any issue, and #17321 appears only on line 1, so the merge will ⛔ not close this card. That is correct, because ruling item 2 is only half delivered (below).
Scope: 8 files, +1062/−38, all inside the claimed surface: conversions/registry.ts + a new test, ui/filter-rule-array.ts + its guidance test, the two ADR-0087 semantic entries whose at-rest prose this made false, the generated migrations/registry.ts, and the changeset. No governed path; 1100 lines.
Diff spot-read by the seat, against ruling 5644018752:
- Item 1: the D2 conversion page-component-filter-record-to-rule-array folds operators through the doors' ONE table (FILTER_OPERATORS + normalizeFilterOperator, with AST infix through parseFilterAST). It is all-or-nothing per filter, and every produced rule must parse at ViewFilterRuleSchema.
- Item 2, pass-through half: any top-level $ key returns unconverted, so ⛔ nothing is flattened.
- Item 3: The seven converged rule-array filter doors refuse the record form with a bare invalid_type — this package's own guidance convention is not applied at any of them #17320 is cross-referenced; only the combinator case's worked rewrite changed (it had prescribed a field literally named $or).
- Item 4: Clause-②: no.
Tests: the new file has 49 cases (exact rule arrays per shape and door, byte-identical pass-through, identity + idempotence, per-operator losslessness, door accept/refuse control). The ablation (the output validation removed) went 3 red and was restored clean. The spec suite passed 541 files; consumers metadata-core 25/25 and lint 107/107.
CI at this head: 13 success / 3 skipped / 16 in progress / 0 failing (an honest in-progress reading).

Contract review: owed on the path limb. An isolated at-tier reviewer is running. needs:contract-review is hung on PR #20175 in the same act. ⛔ Not readied and ⛔ not enqueued before a same-shape PASS on this head.

Deviations

Part of instead of the dispatch's Fixes: accepted, because it is the dispatch's own stop order.
12 doors instead of the named entry's 5: measured and pinned. The at-tier review judges whether that is within the ruling's 「family」.
origin/main was merged once before verification.

The stopped half of ruling item 2, answered by the seat. This is a sequencing question, not a decision owed by the maintainer. The dev's option B (a new card, with this PR switched to Fixes) is ⛔ not taken. The partial-landing route applies: when PR #20175 merges, this card is released in the same act, back to pm:queue with pm:retriage. The Release: line names what landed and the remainder's four sites for triage to route across lanes:

packages/spec/src/conversions/{types,apply,index}.ts has no TODO channel.
packages/runtime/src/dispatcher-error-vocabulary.ts needs a foreign-vocabulary row for a new code.
packages/metadata-protocol/src/protocol.ts: convertStoredItemDetailed forwards only onNotice, and migrateStoredMetadata marks a TODO-only row canonical.
packages/metadata-protocol/src/stored-migration.ts: the report has no TODO field or section.

Findings, one line each

The pending changeset .changeset/19514-… says re-saving object-grid.defaultFilters stays refused, which is now true only for the forms left as stored → Acceptance notes: both changesets ship in the same release, and feat(spec): a stored record-form filter at the rule-array doors converts to the rule array where lossless — D2 page-component-filter-record-to-rule-array (#17321) #20175's own changeset states the new behaviour.
A $-prefixed non-combinator key (for example $text) is rendered by the worked rewrite as a field. Reach is unmeasured → Acceptance notes, not filed.
$null / $exists operator objects are not converted (value-dependent, and not in the one table) → Acceptance notes; they join the TODO list once the remainder lands.

os-dev-report
{
"issue": 17321,
"round": "PATCH round 1 (after contract review 5856719041 PASS at 7459968; PM patch-round order)",
"status": "done",
"branch": "claude/issue-17321-conversion-todo-channel",
"pr": "#20244",
"new_head": "bb0a8300d",
"commits_this_round": [
"4a86b39bd merge origin/main (fdb2669) via scripts/pm/os-regen-merge.sh, no conflicts, no rebase, no force-push",
"26e697044 fix(spec): the D3 entry states the per-door refusal and the TODO listing (plus the defaultFilters leg of the MEASURED test)",
"bb0a8300d chore(spec): regenerate the migration registry mirror (gen:migration-registry; staged diff = the entry text only, +8/-4)"
],
"session": "session_01CiCTczDo7tGhafXjf61dUJ (subagent = parent's)",
"premise_still_valid": true,
"summary": "The review's non-blocking D3 finding is corrected in this PR. The element-data-source D3 entry, and its generated mirror (regenerated, not hand-edited), no longer carry the two false sentences. It now states the per-door refusal (binding refused on save; the properties doors only as the advisory component-props finding) and the TODO listing, matching the changeset twin. The existing MEASURED test gains the properties.defaultFilters door, so every door the corrected text names is measured, not just properties.filter. The tree grep found no other copy of either phrase. It did find the same false claim, reworded, in the sibling D3 entry for object-grid defaultFilters, which is outside this round's surface: reported, not touched. The branch was merged with current main through os-regen-merge.sh. All 15 generated artefacts are current, the suites are green, the gates reconcile, and CI converged with every required check green. Check Changeset is red only for the deliberate correction, naming exactly the #20175 changeset.",
"corrected_sentences_verbatim": {
"removed_1": "os migrate meta --stored does not list these rows yet: a row the conversion leaves as stored reports there as already on protocol.",
"removed_2": "(… by its own contract) and is refused at the filter door on its next save — for a combinator record, a refusal that names the combinator and says why no rule spells it.",
"now_reads (entry reason, from 'Such a row keeps loading unchanged' to the end; the mirror carries the same string)": "Such a row keeps loading unchanged (applyConversionsToStoredItem replays the chain without validating, by its own contract), and its filter door refuses the form: at dataSource.filter on the page's next save; at a block's properties.filter — like properties.defaultFilters, a key of the open properties bag — only as the component-props gate's advisory finding (os validate), since a re-save through the metadata API is not refused there. For a combinator record that refusal names the combinator and says why no rule spells it. os migrate meta --stored lists each filter left as stored as a TODO under its row, naming the block and what blocks the rewrite; a row whose only finding is such a TODO is reported skipped, and the run's exit code does not change for it."
},
"grep_other_copies": "git grep over the whole tree at bb0a830. 'does not list these rows yet' has 2 hits, both removed here (the entry and its mirror); now 0. 'refused at the filter door on its next save' has 2 hits, both removed here; now 0. Wider variants: no hit in content/docs, docs/protocol-upgrade-guide.md, skills/** or apps/docs. The other 'refused on its next save' hits are other families' own doors (sort, the value path, joined reports, the view overlay), not this claim. One same-claim reword outside the surface: packages/spec/src/migrations/entries/semantic/18.object-grid-default-filters-rule-array.ts ('RE-SAVING such a node is refused at the defaultFilters path', plus its mirror); reported below, not touched. The skills/** tree carries neither phrase, so there is nothing Tier H to report.",
"tests": "At bb0a830. spec conversion trio (page-component-filter-record-to-rule-array, stored, conversions): 3 files / 314 passed (one more than round 0: a conversion main added brings its fixture case). metadata-protocol protocol.stored-migration.test.ts: 30/30, including the MEASURED test's new defaultFilters leg (defaultFilters leftover rewritten, combinator persisted byte-identical). spec full suite in two shards (vitest run --project local --shard=N/2): 274 files / 8312 passed + 2 todo, and 273 files / 7796 passed = 547 files. metadata-protocol full suite: 189 passed + 3 skipped files / 2709 passed + 19 skipped; pnpm typecheck exit 0. spec check:generated: 'All 15 generated artifacts are up to date' after rebuilding spec on the regenerated source; check:migration-registry exit 0 ('src/migrations/registry.ts is current (264 semantic, 214 retired-key, 199 retired-def)'). Sibling-entry assertion after the merge (os-regen-merge step 4.3): 6 entry ids main added, plus this card's, are present once each in the staged registry; git diff --cached --stat origin/main -- packages/spec/src/migrations/registry.ts shows only this entry's 12 lines. runtime / metadata-core not re-run this round: their sources are unchanged since 7459968, apart from what the merge brought, which CI's Test Core covered green.",
"gates_head": "bb0a8300d",
"gates_summary": "dispatch-gates --commands --repo objectstack-ai/objectstack at bb0a830 derived 113 (4 more than round 0: check:authorable-surface, check:migration-registry, check:spec-changes, check:future-spec-major, derived from the D3 entry and registry edit). Result: 110 exit 0, 1 exit 1 = check:empty-changeset (its deliberate-correction class; its one ::error names exactly .changeset/17321-record-filter-d2-conversion.md). 2 NOT-MEASURED: check:dual-build-cjs-loads (exit 3, needs the apps' dist) and check:type-check-debt (passed its prerequisite after the closure builds, then hit the 400 s foreground cap, exit 124: a tsc sweep over all 80 packages; CI's 'Type Check · debt ledger' is success). check:skill-examples and check:lean-entry-closure are exit 0 after building the client-react closure. --ran reconciliation: 113 derived, 111 run, 2 NOT-MEASURED, 0 UNRUN, exit 0. Upstream drift: main moved 6 commits after the merge (to 14ae40b), touching cli.mdx and migrations/registry.ts. A driver-less bare-clone merge-tree of bb0a830 with 14ae40b is clean (exit 0), so GitHub can build the merge ref. Not merged a second time.",
"gates": [
"0 :: node scripts/check-adr-0087-registration.mjs --base origin/main",
"0 :: node scripts/check-adr-0087-registration.mjs --self-test",
"0 :: node scripts/check-changeset-no-major.mjs --base origin/main",
"0 :: node scripts/check-changeset-no-major.mjs --self-test",
"0 :: node scripts/check-ci-filter-parity.mjs",
"0 :: node scripts/check-closing-keyword-parity.mjs",
"0 :: node scripts/check-closing-keyword-parity.mjs --self-test",
"0 :: node scripts/check-comment-mask-adoption.mjs",
"0 :: node scripts/check-comment-mask-adoption.mjs --self-test",
"0 :: node scripts/check-comment-mask-corpus.mjs",
"0 :: node scripts/check-dev-prereqs.mjs --self-test",
"0 :: node scripts/check-doc-frontmatter.mjs",
"0 :: node scripts/check-doc-frontmatter.mjs --self-test",
"0 :: node scripts/check-doc-route-spelling.mjs --advisory",
"0 :: node scripts/check-doc-route-spelling.mjs --self-test",
"0 :: node scripts/check-docs-section-name.mjs",
"0 :: node scripts/check-docs-section-name.mjs --self-test",
"1 :: node scripts/check-empty-changeset.mjs --base origin/main",
"0 :: node scripts/check-empty-changeset.mjs --self-test",
"0 :: node scripts/check-keyed-text-bounds.mjs",
"0 :: node scripts/check-keyed-text-bounds.mjs --self-test",
"0 :: node scripts/check-platform-object-tenancy-census.mjs",
"0 :: node scripts/check-platform-object-tenancy-census.mjs --self-test",
"0 :: node scripts/check-plugin-teardown-shape.mjs",
"0 :: node scripts/check-plugin-teardown-shape.mjs --self-test",
"0 :: node scripts/check-registry-log-declared.mjs",
"0 :: node scripts/check-registry-log-declared.mjs --self-test",
"0 :: node scripts/check-rest-log-spy-declared.mjs",
"0 :: node scripts/check-rest-log-spy-declared.mjs --self-test",
"0 :: node scripts/check-section-landing-index.mjs",
"0 :: node scripts/check-section-landing-index.mjs --self-test",
"0 :: node scripts/check-spec-docblock-symbol-anchors.mjs",
"0 :: node scripts/check-spec-docblock-symbol-anchors.mjs --self-test",
"0 :: node scripts/check-system-context-census.mjs",
"0 :: node scripts/check-system-context-census.mjs --self-test",
"0 :: node scripts/check-undeclared-dep-imports.mjs",
"0 :: node scripts/check-undeclared-dep-imports.mjs --self-test",
"0 :: node scripts/docs-audit/check-affected-docs.mjs",
"0 :: node scripts/docs-audit/check-drift-comment.mjs",
"0 :: node scripts/pm/release-rehearsal-clone.mjs --self-test",
"0 :: pnpm --filter @objectstack/lint run check:doc-formula-expressions",
"0 :: pnpm --filter @objectstack/lint run check:doc-security-posture",
"0 :: pnpm --filter @objectstack/spec run check:api-surface",
"0 :: pnpm --filter @objectstack/spec run check:authorable-surface",
"0 :: pnpm --filter @objectstack/spec run check:browser-reachable-entries",
"0 :: pnpm --filter @objectstack/spec run check:docs",
"0 :: pnpm --filter @objectstack/spec run check:dual-source-exports",
"0 :: pnpm --filter @objectstack/spec run check:duration-unit-keys",
"0 :: pnpm --filter @objectstack/spec run check:empty-state",
"0 :: pnpm --filter @objectstack/spec run check:entry-nameability",
"0 :: pnpm --filter @objectstack/spec run check:export-origins",
"0 :: pnpm --filter @objectstack/spec run check:exported-any",
"0 :: pnpm --filter @objectstack/spec run check:generated",
"0 :: pnpm --filter @objectstack/spec run check:liveness",
"0 :: pnpm --filter @objectstack/spec run check:llms-txt",
"0 :: pnpm --filter @objectstack/spec run check:migration-registry",
"0 :: pnpm --filter @objectstack/spec run check:objectui-pin-citations",
"0 :: pnpm --filter @objectstack/spec run check:skill-examples",
"0 :: pnpm --filter @objectstack/spec run check:skill-refs",
"0 :: pnpm --filter @objectstack/spec run check:spec-changes",
"0 :: pnpm --filter @objectstack/spec run check:strictness-ledger",
"0 :: pnpm --filter @objectstack/spec run check:upgrade-guide",
"0 :: pnpm --filter @objectstack/spec run check:variant-docs",
"0 :: pnpm --filter @objectstack/spec run check:yaml-examples",
"0 :: pnpm check:changeset-gate-self-tests",
"0 :: pnpm check:cli-examples-parity",
"0 :: pnpm check:corpus-claim-drift",
"0 :: pnpm check:cross-package-test-inputs",
"0 :: pnpm check:dispatcher-error-vocabulary",
"0 :: pnpm check:doc-anchors",
"0 :: pnpm check:doc-authoring",
"0 :: pnpm check:docs-audit-scope",
"0 :: pnpm check:docs-redirects",
"0 :: pnpm check:docs-single-h1",
"0 :: pnpm check:docs-spec-enumerations",
"0 :: pnpm check:docs-transcript-drift",
"0 :: pnpm check:driver-memory-census",
"0 :: pnpm check:dts-closure",
"3 :: pnpm check:dual-build-cjs-loads",
"0 :: pnpm check:durability-log-level",
"0 :: pnpm check:engine-double-contract",
"0 :: pnpm check:filter-alias-parity",
"0 :: pnpm check:future-spec-major",
"0 :: pnpm check:gitlink-declared",
"0 :: pnpm check:issue-citations",
"0 :: pnpm check:lean-entry-closure",
"0 :: pnpm check:logger-receiver-detach",
"0 :: pnpm check:merge-driver",
"0 :: pnpm check:nul-bytes",
"0 :: pnpm check:objectql-double-limit",
"0 :: pnpm check:objectui-changeset",
"0 :: pnpm check:org-identifier",
"0 :: pnpm check:page-declaration-shape",
"0 :: pnpm check:pm-changeset-deadline-census",
"0 :: pnpm check:pm-prior-rulings",
"0 :: pnpm check:pm-widening-tells",
"0 :: pnpm check:published-files",
"0 :: pnpm check:published-readme-links",
"0 :: pnpm check:query-options-erasure",
"0 :: pnpm check:react-page-adapter-contract",
"0 :: pnpm check:refd-timer-probe",
"0 :: pnpm check:role-word",
"0 :: pnpm check:skill-identifier-liveness",
"0 :: pnpm check:slot-lookup",
"0 :: pnpm check:sourcemap-no-sources-content",
"0 :: pnpm check:spec-parsed-alias",
"0 :: pnpm check:test-source-alias",
"0 :: pnpm check:tier-file-adoption",
"0 :: pnpm check:type-check-coverage",
"NOT-MEASURED pnpm check:type-check-debt :: repo-wide tsc over all 80 packages under a 6144 MB heap; passed its prerequisite, then hit the 400 s foreground timeout on a shared box (exit 124) — a whole-repo sweep, left to CI",
"0 :: pnpm check:vendor-version-stamps",
"0 :: pnpm check:watch-hint-literal",
"0 :: pnpm check:where-matcher"
],
"ci": "Check-runs on bb0a830, read in the foreground and already converged: 35 total, 32 success, 2 skipped (Console Pin Gate; Packed-tarball smoke, opt-in), 1 failure = Check Changeset. Its annotations: one file-scoped failure naming exactly .changeset/17321-record-filter-d2-conversion.md (the empty-changeset foreign-changeset rule, deliberate-correction class), one generic 'Process completed with exit code 1', and one runner notice. Every required context is success: Lint & Repo Gates, TypeScript Type Check (+ workspace / consumer gates / source gates / debt ledger), Test Core (+6 shards), Dogfood Regression Gate (+3 shards), Build Core, Temporal Conformance (live PG + MySQL), Governed Surface Queue Guard. Also success: Spec property liveness, Build Docs, Dogfood Verify CLI, the claim/closing guards.",
"line_budget": "+1016 / -135 = 1151 changed lines vs merge base fdb2669 (19 files), under the 5,000-line threshold; no governed-surface path. This round alone, excluding the merge: 3 files, +26 / -9.",
"files_changed": [
".changeset/17321-conversion-todo-channel.md",
".changeset/17321-record-filter-d2-conversion.md",
"content/docs/deployment/cli.mdx",
"packages/metadata-protocol/src/index.ts",
"packages/metadata-protocol/src/protocol.stored-migration.test.ts",
"packages/metadata-protocol/src/protocol.ts",
"packages/metadata-protocol/src/stored-migration.ts",
"packages/runtime/src/dispatcher-error-vocabulary.ts",
"packages/spec/api-surface/root.json",
"packages/spec/export-origins/root.json",
"packages/spec/src/conversions/apply.ts",
"packages/spec/src/conversions/index.ts",
"packages/spec/src/conversions/page-component-filter-record-to-rule-array.test.ts",
"packages/spec/src/conversions/registry.ts",
"packages/spec/src/conversions/stored.test.ts",
"packages/spec/src/conversions/stored.ts",
"packages/spec/src/conversions/types.ts",
"packages/spec/src/migrations/entries/semantic/18.element-data-source-and-object-block-filter-rule-array.ts",
"packages/spec/src/migrations/registry.ts"
],
"files_changed_this_round": [
"packages/spec/src/migrations/entries/semantic/18.element-data-source-and-object-block-filter-rule-array.ts",
"packages/spec/src/migrations/registry.ts (generated)",
"packages/metadata-protocol/src/protocol.stored-migration.test.ts"
],
"deviations": [
"PR body not updated (order step 5). My agent definition says a dev writes the PR body once, in the pr_create stroke, never PATCHes it, and names later changes in the report for the seat to write. This round's write budget also lists no PR edit. That definition outranks the dispatch, so this is a conflict: the replacement text is in pr_body_patch below, ready to paste. Tests and gate heads in the existing body also date from 7459968; the figures for bb0a830 are in this report.",
"protocol.stored-migration.test.ts edited beyond the order's two text edits: the MEASURED test gains the properties.defaultFilters door. The corrected entry text and the round-0 changeset both say a defaultFilters re-save is not refused, but round 0 had measured only properties.filter, so that claim was unmeasured until now.",
"The entry's corrected text names properties.defaultFilters although that key is its sibling entry's. I followed the order's per-door wording ('at properties.filter / properties.defaultFilters'), phrased as 'like properties.defaultFilters, a key of the open properties bag', so the entry's own scope statement stays true.",
"One spec full-suite attempt died at the outer 595 s cap seconds after acquiring the lock, because most of the budget went to queue wait. No orphan process was left. I re-ran it as two shards with OS_VERIFY_LOCK_WAIT=240.",
"Not merged a second time after main moved 6 more commits. The driver-less merge-tree probe is clean, and the order asked for one merge."
],
"pr_body_patch": {
"where": "Replace the whole '## Acceptance notes' section of PR #20244's body with the text below. In '## Scope beyond the claimed file surface (declared)', add one bullet: '- The D3 entry 18.element-data-source-and-object-block-filter-rule-array.ts and its generated mirror in migrations/registry.ts: two sentences, ordered in the patch round after review 5856719041.' The existing session-URL footer stays last.",
"acceptance_notes_replacement": "## Acceptance notes\n\n- D3 entry corrected in this PR (patch round, ordered after review 5856719041). packages/spec/src/migrations/entries/semantic/18.element-data-source-and-object-block-filter-rule-array.ts and its generated mirror in packages/spec/src/migrations/registry.ts (regenerated with gen:migration-registry, never hand-edited) no longer say "os migrate meta --stored does not list these rows yet" (made false by this PR) or "is refused at the filter door on its next save" (measured false for the properties doors). They now state the per-door refusal and the TODO listing, matching the changeset twin. The measured test now drives all three doors: dataSource.filter leftover gives failed, while properties.filter and properties.defaultFilters leftovers give rewritten, with the combinator persisted byte-identical.\n- Noted for the PM: the CLI closing line (packages/cli/src/commands/migrate/meta.ts, the clean && !apply branch) still prints "Stored metadata is already on protocol N — nothing to rewrite" under the TODO block. It is domain:cli's, and is already false for every existing skip class; the fix is one condition.\n- Noted for the PM: retiring the inline-row decline once .objectui-sha carries objectui#10767 (upstream 17b323e5ca; the pin f8a9d0fb05 does not carry it). This PR keeps the decline and words its TODO as the pinned renderer's limit.\n- Noted for the PM: the twin D3 entry 18.object-grid-default-filters-rule-array.ts (and its mirror) still says "RE-SAVING such a node is refused at the defaultFilters path". That is measured false in this PR's test (the defaultFilters leftover is rewritten), but the file is outside this round's surface, so it is not touched here.\n"
},
"mcp_calls": "0",
"api_writes": "1 REST write this round: this os-dev-report comment, POST /repos//issues/17321/comments, via the fleet relay (repository_dispatch). Plus git pushes of 26e6970 and bb0a830 (not REST). No PR edit, no label write, no readiness change. Reads: GET issue 17321 comments, PR 20244, issue comment 5856719041, the check-runs of bb0a830 and the Check Changeset run's annotations.",
"open_questions": [],
"out_of_scope_findings": [
"class: b · reach: exception: release-text — packages/spec/src/migrations/entries/semantic/18.object-grid-default-filters-rule-array.ts (reason, 'RE-SAVING such a node is refused at the defaultFilters path, with the same conversion table the filter door gives, computed from the author\'s own keys') and its generated mirror in migrations/registry.ts. This is the same family and the same false claim the patch round just corrected in the element-data-source twin. Evidence: protocol.stored-migration.test.ts on #20244 at bb0a830, MEASURED test, page 'deal_grid': a stored object-grid whose defaultFilters is a combinator record re-saves on --apply (rewritten, combinator persisted byte-identical). The fix is one sentence plus gen:migration-registry and can ride #20244 if the PM orders it. Suggested text: 'and its door refuses the form only as the component-props gate\'s advisory finding (os validate) — a re-save through the metadata API is not refused there — with the same conversion table the filter door gives, computed from the author\'s own keys.' · dedupe words: 'RE-SAVING such a node is refused at the defaultFilters path', 'object-grid defaultFilters re-save', 'properties bag advisory refusal'",
"carrier: domain:cli seat · noted for the PM, not filed — the CLI closing line 'Stored metadata is already on protocol N — nothing to rewrite' (packages/cli/src/commands/migrate/meta.ts, clean && !apply branch) prints under the ☐ TODO block. Already false for every existing skip class; the fix is one condition.",
"carrier: PM · noted for the PM, not filed — retire the rendersInlineRows decline once .objectui-sha carries objectui#10767 (17b323e5ca); the pin f8a9d0fb05 does not."
]
}

Generated by Claude Code

Claim `5855123916`: file surface extended for a second patch round on PR #20244 (2026-09-27T16:09Z)

domain:spec seat 4 (session_01CiCTczDo7tGhafXjf61dUJ). ⛔ Not a new claim: branch, worktree, assignee and Clause-②: yes are unchanged.

Why. The patch round (report 5857520707, head bb0a8300d) corrected the element-data-source D3 entry. It then found the same false sentence in the twin D3 entry packages/spec/src/migrations/entries/semantic/18.object-grid-default-filters-rule-array.ts: 「RE-SAVING such a node is refused at the defaultFilters path」. This PR's own test measures that claim false: a stored object-grid whose defaultFilters is a combinator record re-saves as rewritten, with the combinator persisted. The object-grid defaultFilters door is one of the doors this card's conversion (page-component-filter-record-to-rule-array) covers. So this is the same family and the same correction, and it rides this PR instead of a new card.

Added to the surface: that entry file's reason text (one sentence), and its generated mirror in packages/spec/src/migrations/registry.ts, regenerated with gen:migration-registry and never hand-edited.

os-dev-report
{
"issue": 17321,
"round": "PATCH round 3 (claim amendment 5857535828: the object-grid defaultFilters D3 entry's one sentence + its generated mirror)",
"status": "done",
"branch": "claude/issue-17321-conversion-todo-channel",
"pr": "#20244",
"new_head": "755acf300",
"commits_this_round": [
"42ecba319 merge origin/main (a9fb83e) via scripts/pm/os-regen-merge.sh; main had moved under migrations/registry.ts (2 commits: cfc3bcf #20247, a91d12a #20245); no conflicts, no rebase, no force-push",
"847f1ab57 fix(spec): the defaultFilters D3 entry states where its door refuses, not a re-save refusal",
"755acf300 chore(spec): regenerate the migration registry mirror (gen:migration-registry; staged diff = this sentence only, +6/-3)"
],
"session": "session_01CiCTczDo7tGhafXjf61dUJ (subagent = parent's)",
"premise_still_valid": true,
"summary": "The false sentence in 18.object-grid-default-filters-rule-array.ts ('RE-SAVING such a node is refused at the defaultFilters path') is replaced, and its mirror in migrations/registry.ts is regenerated. Before writing, I checked each clause of the proposed replacement against the code at this head. Two clauses were not true as proposed and are written true instead. (1) '(os validate)' was too narrow: the component-props rule runs on os validate, os build and os lint (commands ALL = AUTHORING_COMMANDS, tier advisory, surfaces CLI_ONLY), and never on the runtime publish gate. (2) 'with the same conversion table the filter door gives, computed from the author's own keys' holds only for a record-form leftover. The entry's own leftover list includes the bare string or number this key took and AST forms, and the door answers those with the schema's plain type refusal (measured below). The clause 'a re-save through the metadata API is not refused there' was re-measured at this head after the merge brought #20247 (a save-door filter change) and still holds. The branch was merged with current main first, because main had moved under the generated registry.",
"sentence_verbatim": {
"removed": "and RE-SAVING such a node is refused at the defaultFilters path, with the same conversion table the filter door gives, computed from the author's own keys.",
"now_reads (reason, from 'What it cannot map losslessly' to the end; the mirror carries the same string)": "What it cannot map losslessly is left exactly as stored and keeps rendering as it does today — a combinator, a null value, an operator the rule vocabulary does not spell, the bare string or number this key also took, and any filter on a grid whose rows are inline (data with provider value, or staticData), for the reason its sibling gives — and its door refuses such a value only as the component-props gate's advisory finding (os validate, os build, os lint), since a re-save through the metadata API is not refused there: a record form with the message the filter door gives, a worked rewrite computed from the author's own keys and a pointer to this entry's conversion table, and a bare string or number or an AST tuple array with the schema's plain type refusal. ADR-0049 / ADR-0087."
},
"clause_check": [
"'its door refuses such a value': TRUE. component.zod.ts ComponentPropsMap['object-grid'].defaultFilters is z.array(ViewFilterRuleSchema, { error: ruleArrayFilterError({ migration: 'object-grid-default-filters-rule-array' }) }), and every leftover kind the entry lists fails that parse (probe below).",
"'only as the component-props gate's advisory finding': TRUE. lint authoring-rules.ts validateComponentProps is tier 'advisory', commands ALL (= AUTHORING_COMMANDS ['validate','build','lint']), surfaces CLI_ONLY; the runtime publish gate runs only rules whose surfaces include 'runtime-publish'. '(os validate)' was CORRECTED to '(os validate, os build, os lint)'.",
"'a re-save through the metadata API is not refused there': TRUE, re-measured at 755acf3. protocol.stored-migration.test.ts 'MEASURED — the write path judges the two door kinds differently', page 'deal_grid': a stored object-grid whose defaultFilters is a combinator record is rewritten on --apply, with the combinator persisted byte-identical (30/30 at this head, after the #20247 merge).",
"'with the same conversion table the filter door gives, computed from the author's own keys': TRUE ONLY FOR A RECORD FORM, so CORRECTED to name both answers. Probe of ComponentPropsMap['object-grid'].safeParse({ defaultFilters }) at this head (tsx on source): the record combinator, the null value and the $null operator are each refused at defaultFilters with the ruleArrayFilterError message, which carries the worked rewrite and 'Full conversion table: migration object-grid-default-filters-rule-array'. The bare string gets 'Invalid input: expected array, received string' and the bare number 'expected array, received number', with no rewrite and no table. The AST [['name','like','%a%']] is refused at defaultFilters.0 with 'expected object, received array', also with no rewrite and no table. The lint gate passes the message through (describeIssue returns issue.message). 'the same message the filter door gives' is literal: one ruleArrayFilterError map, with its surface naming the defaultFilters fallback; the existing pin component-object-grid-default-filters.pin.test.ts asserts the record message (lines 85-89) and that element-level issues carry no migration pointer (line 133)."
],
"tests": "At 755acf3, one verify-lock acquisition (held 68 s). metadata-protocol closure build: turbo 12/12 successful. spec src/migrations (3 files) + ui/component-object-grid-default-filters.pin.test.ts + ui/filter-rule-array-guidance.test.ts: 5 files / 201 passed. Conversion trio: 3 files / 314 passed. metadata-protocol protocol.stored-migration.test.ts: 30/30, including the MEASURED defaultFilters leg. Exit codes on disk: closure-build 0, spec-migrations+pins 0, conversion-trio 0, stored-migration 0.",
"gates_head": "755acf300",
"gates_named_by_order": [
"0 :: pnpm --filter @objectstack/spec run check:migration-registry @ 755acf3",
"0 :: pnpm check:issue-citations && node scripts/check-issue-citations.mjs @ 755acf3",
"0 :: pnpm check:doc-authoring @ 755acf3",
"0 :: pnpm check:nul-bytes @ 755acf3",
"0 :: pnpm --filter @objectstack/spec run check:generated @ 755acf3",
"0 :: pnpm --filter @objectstack/spec run check:spec-changes @ 755acf3",
"0 :: pnpm --filter @objectstack/spec run check:upgrade-guide @ 755acf3",
"0 :: pnpm --filter @objectstack/spec run check:api-surface @ 755acf3"
],
"gates_summary": "Named by the order, all exit 0 at 755acf3: check:migration-registry ('registry.ts is current (266 semantic, 214 retired-key, 199 retired-def)'); pnpm check:issue-citations && node scripts/check-issue-citations.mjs, whose judging pass reports 'citations judged: 4 across 12 file(s) — 3 resolves, 1 cross-repo-unjudged … every citation this change adds resolves (or is a declared cross-repo reference)'; check:generated ('All 15 generated artifacts are up to date', after building spec at this head). Also exit 0: check:spec-changes, check:upgrade-guide, check:api-surface, check:doc-authoring, check:nul-bytes. The full derived set (dispatch-gates --commands --repo objectstack-ai/objectstack at 755acf3, the same 113 families as round 2) was also run. Result: 108 exit 0, 1 exit 1 = check:empty-changeset (the deliberate correction; its one ::error names exactly .changeset/17321-record-filter-d2-conversion.md), and 4 exit 3 = NOT MEASURED (check:skill-examples needs the client-react closure; check:lean-entry-closure needs objectql dist; check:dual-build-cjs-loads and check:type-check-debt need every package's dist). check:skill-examples and check:lean-entry-closure were exit 0 at bb0a830 last round, and this round changes neither's inputs beyond main's merge. --ran reconciliation: 113 derived, 109 run, 4 NOT-MEASURED (derived from exit 3), 0 UNRUN, exit 0.",
"gates": [
"0 :: node scripts/check-adr-0087-registration.mjs --base origin/main",
"0 :: node scripts/check-adr-0087-registration.mjs --self-test",
"0 :: node scripts/check-changeset-no-major.mjs --base origin/main",
"0 :: node scripts/check-changeset-no-major.mjs --self-test",
"0 :: node scripts/check-ci-filter-parity.mjs",
"0 :: node scripts/check-closing-keyword-parity.mjs",
"0 :: node scripts/check-closing-keyword-parity.mjs --self-test",
"0 :: node scripts/check-comment-mask-adoption.mjs",
"0 :: node scripts/check-comment-mask-adoption.mjs --self-test",
"0 :: node scripts/check-comment-mask-corpus.mjs",
"0 :: node scripts/check-dev-prereqs.mjs --self-test",
"0 :: node scripts/check-doc-frontmatter.mjs",
"0 :: node scripts/check-doc-frontmatter.mjs --self-test",
"0 :: node scripts/check-doc-route-spelling.mjs --advisory",
"0 :: node scripts/check-doc-route-spelling.mjs --self-test",
"0 :: node scripts/check-docs-section-name.mjs",
"0 :: node scripts/check-docs-section-name.mjs --self-test",
"1 :: node scripts/check-empty-changeset.mjs --base origin/main",
"0 :: node scripts/check-empty-changeset.mjs --self-test",
"0 :: node scripts/check-keyed-text-bounds.mjs",
"0 :: node scripts/check-keyed-text-bounds.mjs --self-test",
"0 :: node scripts/check-platform-object-tenancy-census.mjs",
"0 :: node scripts/check-platform-object-tenancy-census.mjs --self-test",
"0 :: node scripts/check-plugin-teardown-shape.mjs",
"0 :: node scripts/check-plugin-teardown-shape.mjs --self-test",
"0 :: node scripts/check-registry-log-declared.mjs",
"0 :: node scripts/check-registry-log-declared.mjs --self-test",
"0 :: node scripts/check-rest-log-spy-declared.mjs",
"0 :: node scripts/check-rest-log-spy-declared.mjs --self-test",
"0 :: node scripts/check-section-landing-index.mjs",
"0 :: node scripts/check-section-landing-index.mjs --self-test",
"0 :: node scripts/check-spec-docblock-symbol-anchors.mjs",
"0 :: node scripts/check-spec-docblock-symbol-anchors.mjs --self-test",
"0 :: node scripts/check-system-context-census.mjs",
"0 :: node scripts/check-system-context-census.mjs --self-test",
"0 :: node scripts/check-undeclared-dep-imports.mjs",
"0 :: node scripts/check-undeclared-dep-imports.mjs --self-test",
"0 :: node scripts/docs-audit/check-affected-docs.mjs",
"0 :: node scripts/docs-audit/check-drift-comment.mjs",
"0 :: node scripts/pm/release-rehearsal-clone.mjs --self-test",
"0 :: pnpm --filter @objectstack/lint run check:doc-formula-expressions",
"0 :: pnpm --filter @objectstack/lint run check:doc-security-posture",
"0 :: pnpm --filter @objectstack/spec run check:api-surface",
"0 :: pnpm --filter @objectstack/spec run check:authorable-surface",
"0 :: pnpm --filter @objectstack/spec run check:browser-reachable-entries",
"0 :: pnpm --filter @objectstack/spec run check:docs",
"0 :: pnpm --filter @objectstack/spec run check:dual-source-exports",
"0 :: pnpm --filter @objectstack/spec run check:duration-unit-keys",
"0 :: pnpm --filter @objectstack/spec run check:empty-state",
"0 :: pnpm --filter @objectstack/spec run check:entry-nameability",
"0 :: pnpm --filter @objectstack/spec run check:export-origins",
"0 :: pnpm --filter @objectstack/spec run check:exported-any",
"0 :: pnpm --filter @objectstack/spec run check:generated",
"0 :: pnpm --filter @objectstack/spec run check:liveness",
"0 :: pnpm --filter @objectstack/spec run check:llms-txt",
"0 :: pnpm --filter @objectstack/spec run check:migration-registry",
"0 :: pnpm --filter @objectstack/spec run check:objectui-pin-citations",
"3 :: pnpm --filter @objectstack/spec run check:skill-examples",
"0 :: pnpm --filter @objectstack/spec run check:skill-refs",
"0 :: pnpm --filter @objectstack/spec run check:spec-changes",
"0 :: pnpm --filter @objectstack/spec run check:strictness-ledger",
"0 :: pnpm --filter @objectstack/spec run check:upgrade-guide",
"0 :: pnpm --filter @objectstack/spec run check:variant-docs",
"0 :: pnpm --filter @objectstack/spec run check:yaml-examples",
"0 :: pnpm check:changeset-gate-self-tests",
"0 :: pnpm check:cli-examples-parity",
"0 :: pnpm check:corpus-claim-drift",
"0 :: pnpm check:cross-package-test-inputs",
"0 :: pnpm check:dispatcher-error-vocabulary",
"0 :: pnpm check:doc-anchors",
"0 :: pnpm check:doc-authoring",
"0 :: pnpm check:docs-audit-scope",
"0 :: pnpm check:docs-redirects",
"0 :: pnpm check:docs-single-h1",
"0 :: pnpm check:docs-spec-enumerations",
"0 :: pnpm check:docs-transcript-drift",
"0 :: pnpm check:driver-memory-census",
"0 :: pnpm check:dts-closure",
"3 :: pnpm check:dual-build-cjs-loads",
"0 :: pnpm check:durability-log-level",
"0 :: pnpm check:engine-double-contract",
"0 :: pnpm check:filter-alias-parity",
"0 :: pnpm check:future-spec-major",
"0 :: pnpm check:gitlink-declared",
"0 :: pnpm check:issue-citations",
"3 :: pnpm check:lean-entry-closure",
"0 :: pnpm check:logger-receiver-detach",
"0 :: pnpm check:merge-driver",
"0 :: pnpm check:nul-bytes",
"0 :: pnpm check:objectql-double-limit",
"0 :: pnpm check:objectui-changeset",
"0 :: pnpm check:org-identifier",
"0 :: pnpm check:page-declaration-shape",
"0 :: pnpm check:pm-changeset-deadline-census",
"0 :: pnpm check:pm-prior-rulings",
"0 :: pnpm check:pm-widening-tells",
"0 :: pnpm check:published-files",
"0 :: pnpm check:published-readme-links",
"0 :: pnpm check:query-options-erasure",
"0 :: pnpm check:react-page-adapter-contract",
"0 :: pnpm check:refd-timer-probe",
"0 :: pnpm check:role-word",
"0 :: pnpm check:skill-identifier-liveness",
"0 :: pnpm check:slot-lookup",
"0 :: pnpm check:sourcemap-no-sources-content",
"0 :: pnpm check:spec-parsed-alias",
"0 :: pnpm check:test-source-alias",
"0 :: pnpm check:tier-file-adoption",
"0 :: pnpm check:type-check-coverage",
"3 :: pnpm check:type-check-debt",
"0 :: pnpm check:vendor-version-stamps",
"0 :: pnpm check:watch-hint-literal",
"0 :: pnpm check:where-matcher"
],
"ci": "Check-runs on 755acf3 had already converged at one read (no wait ordered this round): 35 total, 32 success, 2 skipped (Console Pin Gate; Packed-tarball smoke, opt-in), 1 failure = Check Changeset. Its file annotation is exactly .changeset/17321-record-filter-d2-conversion.md, plus the generic exit line and a runner notice. Every required context is success.",
"line_budget": "+1028 / -141 = 1169 changed lines vs merge base a9fb83e (20 files), under the 5,000-line threshold; no governed-surface path. This round alone, excluding the merge: 2 files, +12 / -6.",
"files_changed_this_round": [
"packages/spec/src/migrations/entries/semantic/18.object-grid-default-filters-rule-array.ts (reason: one sentence)",
"packages/spec/src/migrations/registry.ts (generated mirror)"
],
"deviations": [
"The replacement sentence differs from the proposed one in two clauses (see clause_check): '(os validate)' became '(os validate, os build, os lint)', and the conversion-table clause is split into the record-form answer and the plain type refusal. The order said to write the true clause where the proposal was not true, and to say so. It is still one sentence, per the amendment.",
"Heavy runs waited on a deep shared-lock queue: four queue-timeouts (exit 99, NOT MEASURED, slot kept), then one acquisition with OS_VERIFY_LOCK_WAIT=330 and a 240 s run cap so wait plus run fit the foreground limit. All four heavy steps ran in that single hold.",
"I ran the full derived gate set in addition to the families the order named."
],
"mcp_calls": "0",
"api_writes": "1 REST write: this os-dev-report comment, POST /repos//issues/17321/comments, via the fleet relay (repository_dispatch). Plus git push of 42ecba3, 847f1ab and 755acf3 (not REST). No PR edit, no label write, no readiness change. Reads: GET issue comment 5857535828, the check-runs of 755acf3 and the Check Changeset annotations.",
"open_questions": [],
"out_of_scope_findings": [
"carrier: PM (this PR, if ordered) · noted, not touched (outside amendment 5857535828) — a precision gap in the round-2 sentence of 18.element-data-source-and-object-block-filter-rule-array.ts: 'os migrate meta --stored lists each filter left as stored as a TODO under its row'. In context 'each filter' means the leftover forms that entry enumerates (combinator, null value, $null/$exists, AST like, array or object comparand, AST group, inline rows), and every one of those is reported. But that entry also records that the four object-* block doors used to take z.unknown(), so a stored bare string or number at properties.filter there is also left as stored, and it is NOT reported: the conversion recognises only record and AST forms (pinned in page-component-filter-record-to-rule-array.test.ts §8, 'a value that is not a legacy form is neither converted nor reported'). Suggested one-word fix: 'lists each such filter left as stored as a TODO under its row', plus gen:migration-registry."
]
}

Generated by Claude Code

os-dev-report
{
"issue": 17321,
"round": "PATCH round 4 (after delta review 5858721368 PASS at 755acf3: merge conflict + precision tightening)",
"status": "done",
"branch": "claude/issue-17321-conversion-todo-channel",
"pr": "#20244",
"new_head": "bafa19790",
"commits_this_round": [
"930fae92e merge origin/main (2dccb7d) via scripts/pm/os-regen-merge.sh: one content conflict (packages/metadata-protocol/src/protocol.ts, the @objectstack/spec import block), resolved keeping both sides; conversions/registry.ts and migrations/registry.ts auto-merged; no rebase, no force-push",
"68294ccbc fix(spec): the element-data-source D3 entry and the pending #20175 note name the unreported non-legacy value and all three props-gate commands",
"bafa19790 chore(spec): regenerate the migration registry mirror (gen:migration-registry; staged diff = this entry's text only, +8/-5)"
],
"session": "session_01CiCTczDo7tGhafXjf61dUJ (subagent = parent's)",
"premise_still_valid": true,
"summary": "The PR is merged with main again and carries the review's precision tightening. The one conflict, the protocol.ts import block, is resolved by keeping main's IObjectQLEngine on the contracts import and this PR's ConversionTodoNotice on the root import. The merge commit's tree equals git's own driver-less three-way merge of its two parents everywhere except that single hunk. The helper's rerun (step 2 against the recorded pre-merge base) took main's side of every generated path main moved, which changed nothing, and check:generated reports all 15 artefacts current. The element-data-source D3 entry and the pending #20175 note get the ordered wording. One clause was not true as ordered and is written true: 'and its row reads as already on protocol' holds only for a row with nothing else, so it reads 'a row carrying nothing else reads as already on protocol'. The mirror was regenerated, and every ordered gate is recorded at bafa197.",
"merge_resolution": {
"conflict": "packages/metadata-protocol/src/protocol.ts only (driver-less merge-tree of 755acf3 x 2dccb7d: exit 1; stages 1/2/3 listed for protocol.ts alone). Base: import type { IPubSub } …/contracts' + import { applyConversionsToStoredItem, type ConversionNotice } from '@objectstack/spec'. Main (#20265, 1207baf) changed line 1 to add IObjectQLEngine; this PR changed line 2 to add ConversionTodoNotice.",
"resolved_hunk_verbatim": "import type { IObjectQLEngine, IPubSub } from '@objectstack/spec/contracts';\nimport { applyConversionsToStoredItem, type ConversionNotice, type ConversionTodoNotice } from '@objectstack/spec';",
"proof": "In a bare --shared probe clone with no merge driver registered (merge.os-regen.driver unset, exit 1), git merge-tree --write-tree 755acf300 2dccb7d49 gave tree dba19f2ca with conflict markers in protocol.ts. git diff dba19f2ca 930fae92e touches 1 file, +1/-6: the conflict-marker block (the seven-less-than-sign line naming 755acf3, the seven-equals-sign separator, the seven-greater-than-sign line naming 2dccb7d, and both sides' two lines) becomes the two lines above. Nothing else differs. So the merge commit's diff against its first parent 755acf3 is exactly main's content since the base, as git's clean three-way merge produces it, plus that one resolution. IObjectQLEngine occurs 3 times in protocol.ts at the head and 3 times on main (the import and its two uses). The os-regen-merge.sh rerun kept the branch's api-surface/root.json and export-origins/root.json (changed by the branch, not by main), took main's side of every generated path main moved (no tree change resulted), and made no extra commit. Sibling assertion: main's one new semantic entry since the base, cel-predicate-one-value-comparand-refused, is present once in the regenerated registry."
},
"precision_edits": {
"entry_now_reads (reason, from 'Such a row keeps loading unchanged' to the end; the mirror carries the same string)": "Such a row keeps loading unchanged (applyConversionsToStoredItem replays the chain without validating, by its own contract), and its filter door refuses the form: at dataSource.filter on the page's next save; at a block's properties.filter — like properties.defaultFilters, a key of the open properties bag — only as the component-props gate's advisory finding (os validate, os build, os lint), since a re-save through the metadata API is not refused there. For a combinator record that refusal names the combinator and says why no rule spells it. os migrate meta --stored lists each such filter as a TODO under its row, naming the block and what blocks the rewrite (a value that is not a record or AST form at all — a bare string or number one of the former z.unknown() doors took — is neither converted nor reported, and a row carrying nothing else reads as already on protocol); a row whose only finding is such a TODO is reported skipped, and the run's exit code does not change for it.",
"changeset_words_changed (git diff --word-diff; the rest of the diff is a whitespace reflow inside the same sentence)": [
"advisory finding (os [-validate)-]{+validate, os build, os lint)+}", "os migrate meta --storedlists each {+such+} filter left as stored as a TODO under its row." ], "clause_check_at_merged_head": [ "'lists each such filter as a TODO under its row, naming the block and what blocks the rewrite': TRUE. Unchanged code: pushTodos nests a TODO under every listed row, and the reason opens 'On thetypeblock …'. At the head, the conversion trio (314) and §8 are green in the 515-test spec run.", "'a value that is not a record or AST form at all … is neither converted nor reported': TRUE. conversions/registry.ts legacyFilterToRuleArray tries recordFilterToRules only when isRecordForm(value) and astFilterToRules only for an array, which returns undefined unless isFilterAST. A bare string or number reachesif (!mapping) return holder;(registry.ts :10877), so there is no emit and no reportTodo. Pinned in §8 control 'a value that is not a legacy form is neither converted nor reported'.", "'a bare string or number one of the formerz.unknown()doors took': TRUE. The same entry records the four object-* block doors as the accept-anythingz.unknown()(entry lines 12, 22, 39).", "'and its row reads as already on protocol' as ORDERED: NOT TRUE in general, so WRITTEN TRUE. protocol.ts migrateStoredMetadata:changed = notices.length > 0, so a row with a convertible filter elsewhere is pending/rewritten; a row with a TODO elsewhere is skipped; only a row with neither is canonicaland counted as on protocol. Written: 'and a row carrying nothing else reads as already on protocol'.", "'(os validate, os build, os lint)': TRUE. lint authoring-rules.ts validateComponentProps has tier advisory, commands ALL = AUTHORING_COMMANDS ['validate','build','lint'] and surfaces CLI_ONLY, the same as in round 3 and as the review measured on all three commands." ] }, "tests": "At bafa19790, all under os-verify-lock. The metadata-protocol closure build (incl. @objectstack/lint, changed by #20265): turbo 12/12. spec src/migrations + ui/component-object-grid-default-filters.pin + ui/filter-rule-array-guidance + the conversion trio: 8 files / 515 passed. metadata-protocol protocol.stored-migration.test.ts + protocol.runtime-authoring-gate.test.ts (the suite of #20265, the merge's protocol.ts change): 2 files / 69 passed. metadata-protocol full suite, run in two shards after a single-run attempt hit my 190 s cap (exit 124, not a verdict): 96 files / 1195 passed, and 93 passed + 3 skipped files / 1530 passed + 19 skipped, for 189 + 3 skipped files and 2725 tests. metadata-protocolpnpm typecheckexit 0.", "heavy_exit_codes": [ "bafa19790 closure-build EXIT=0", "bafa19790 spec-migrations+pins+conversion-trio EXIT=0", "bafa19790 mp stored-migration+runtime-authoring-gate EXIT=0", "bafa19790 mp-full-suite EXIT=124", "bafa19790 mp-typecheck EXIT=0", "bafa19790 mp-full-suite shard1/2 EXIT=0 (96 files, 1195 tests)", "bafa19790 mp-full-suite shard2/2 EXIT=0 (93+3 skipped files, 1530+19 skipped tests)" ], "gates_head": "bafa19790", "gates": [ "0 :: pnpm --filter @objectstack/spec run check:migration-registry @ bafa19790", "0 :: pnpm --filter @objectstack/spec run check:generated @ bafa19790", "0 :: pnpm check:issue-citations && node scripts/check-issue-citations.mjs @ bafa19790", "1 :: node scripts/check-empty-changeset.mjs --base origin/main @ bafa19790", "0 :: pnpm check:nul-bytes @ bafa19790", "0 :: pnpm check:doc-authoring @ bafa19790", "0 :: pnpm check:dispatcher-error-vocabulary @ bafa19790" ], "gates_summary": "Ordered gates at bafa19790: check:migration-registry 0 ('registry.ts is current (267 semantic, 214 retired-key, 199 retired-def)'); check:generated 0 ('All 15 generated artifacts are up to date', after building spec at this head);pnpm check:issue-citations && node scripts/check-issue-citations.mjs0 ('citations judged: 4 across 12 file(s) — 3 resolves, 1 cross-repo-unjudged');node scripts/check-empty-changeset.mjs --base origin/main1, with exactly one ::error,file=.changeset/17321-record-filter-d2-conversion.md(the deliberate correction; no other file). Also 0: check:nul-bytes, check:doc-authoring, check:dispatcher-error-vocabulary. Control-byte grep over this round's three files finds none.", "ci": "One read, no wait ordered: 33 check-runs on bafa19790 so far — 25 success, 2 skipped, 5 in_progress (Test Core 1/3/5/6 of 6, Type Check · workspace), 1 failure = Check Changeset, whose file annotation is exactly.changeset/17321-record-filter-d2-conversion.md(plus the generic exit line and a runner notice). Not converged: in_progress is the honest value. PR: draft, mergeable recomputing (null / unknown) at the read. origin/main has since moved 2 commits to 733822cf8; a driver-less merge-tree of bafa19790 with it is clean (exit 0).", "line_budget": "This round, excluding the merge: 3 files, +21 / -14.", "files_changed_this_round": [ ".changeset/17321-record-filter-d2-conversion.md (two phrases in the ordered sentences; reflow inside that sentence)", "packages/spec/src/migrations/entries/semantic/18.element-data-source-and-object-block-filter-rule-array.ts", "packages/spec/src/migrations/registry.ts (generated)", "packages/metadata-protocol/src/protocol.ts (merge resolution only, in 930fae92e)" ], "deviations": [ "One ordered clause written differently, as the order allows: 'and its row reads as already on protocol' became 'and a row carrying nothing else reads as already on protocol' (see clause_check).", "In the changeset I spelled the three commandsos validate, os build, os lint` with each command in backticks, the note's own style for commands. The D3 entry uses the plain '(os validate, os build, os lint)' the order gave, matching its twin.",
"The changeset's edited sentence is re-wrapped within itself. The word-level diff is exactly the two ordered changes; no other sentence of the note moves."
],
"mcp_calls": "0",
"api_writes": "1 REST write: this os-dev-report comment, POST /repos//issues/17321/comments, via the fleet relay (repository_dispatch). Plus git push of 930fae9, 68294cc and bafa197 (not REST). No PR edit, no label write, no readiness change. Reads: GET review 5858721368, PR 20244, the check-runs of bafa197 and the Check Changeset annotations.",
"open_questions": [],
"out_of_scope_findings": []
}

Generated by Claude Code

ACCEPT — PR #20244 (2026-09-27T20:02Z)

domain:spec seat 4 (session_01CiCTczDo7tGhafXjf61dUJ). This is the review of record for the round claimed in 5855123916. The same claim was widened by amendment 5857535828. The dev reports are 5856538210 (round 1), 5857520707, 5858406062 and 5859219730 (patch rounds 1, 3 and 4).

Checklist, read on GitHub:

PR form: base main. The first line is Fixes #17321, and there is no other closing keyword. The bare line Clause-②: yes matches the diff: new public surface in @objectstack/spec and @objectstack/metadata-protocol. Assignee os-sales.
Scope: 20 files (+1034 / −140). The structured TODO channel of ruling 5644018752 item 2:
- conversions/ types / apply / stored / index, and the page-component-filter-record-to-rule-array pass-through branch;
- metadata-protocol convertStoredItemDetailed / migrateStoredMetadata;
- the stored-migration report;
- the two D3 entries' corrected sentences and their mirrors;
- tests, and the changesets.
At-tier contract review:
- PASS 5856719041 at 7459968a6.
- Delta PASS 5858721368 at 755acf300: both corrected D3 sentences are true.
- Delta PASS 5859338031 at bafa19790e:
  - the merge 930fae92e equals git's own merge of its parents, except the one protocol.ts import resolution, which keeps both sides' names;
  - the tightened element-data-source sentence is true clause by clause, including "a row carrying nothing else reads as already on protocol" against migrateStoredMetadata's outcome logic;
  - the mirror is verbatim;
  - nothing else moved.
Grade: Clause-②: yes; @objectstack/spec minor, @objectstack/metadata-protocol minor.
Deliberate correction: .changeset/17321-record-filter-d2-conversion.md. Every rewritten sentence was judged true at this head (5859338031 ③). The red Check Changeset is carried on the three conditions, recorded on the PR in 5859355312: the gate and cause are named, it is red by design on a pushed branch, and it does not run on merge_group and is not required.
CI at bafa19790e: every required context is success. The single failure is Check Changeset, the carried red above.
Merge: a no-driver merge-tree onto today's main (6a6a17b62) exits 0, and so do merges with PR fix(spec): one D3 entry per major-18 retirement family — the census and the 25 missing entries (#20201) #20255's and PR feat(spec)!: retire the flattened view overlay's owner and hidden keys (ADR-0049) #20286's heads.
Governed surface: NOT governed; 1174 changed lines.

The dev's open questions (5856538210), answered:

Q1 → A. Confirm both corrections of feat(spec): a stored record-form filter at the rule-array doors converts to the rule array where lossless — D2 page-component-filter-record-to-rule-array (#17321) #20175's pending note. The TODO-listing sentence is made true by this PR, and the per-door refusal wording was measured. Restoring the note would republish two false sentences.
Q2 → A. Keep the three declared extras in this PR: cli.mdx (one row), the metadata-protocol index.ts re-export, and the feat(spec): a stored record-form filter at the rule-array doors converts to the rule array where lossless — D2 page-component-filter-record-to-rule-array (#17321) #20175 note. Each is a line-scale edit that keeps a published or pending text true in the same landing as the behaviour. A docs-only split would ship a window in which those texts are false.

Findings, disposed one line each:

The element-data-source D3 entry's 「refused at the filter door on its next save」 was false for the properties doors → fixed in this PR (patch rounds 1 and 4). The object-grid twin was fixed in round 3.
The inline-row decline (rendersInlineRows) protects the pinned renderer, and becomes removable once .objectui-sha carries objectui#10767 → filed spec: retire the inline-row decline in page-component-filter-record-to-rule-array once the objectui pin carries objectui#10767 #20305, pm:on-hold, woken by .objectui-sha.
The CLI's closing line 「Stored metadata is already on protocol N — nothing to rewrite」 prints under a TODO block on a TODO-only os migrate meta --stored run (packages/cli/src/commands/migrate/meta.ts, the clean && !apply branch) → Acceptance notes, carrier the domain:cli lane. Its reach is derived from storedMigrationClean and was not run at the CLI door, so it is not filed. It sits outside this claim.

Next: ready → auto-merge through the relay → merge queue. This card closes on the merge through Fixes.

Landed — PR #20244 → `7b1e4a4871` (2026-09-27T21:12Z)

domain:spec seat 4 (session_01CiCTczDo7tGhafXjf61dUJ). PR #20244 merged through the merge queue on 2026-09-27T20:24:52Z, and this card closed completed through its Fixes line. The deliberate-correction red of Check Changeset was carried on the three conditions recorded in 5859355312. The readings that confirm the landing:

7b1e4a4871 has one parent (f415bcf186), is an ancestor of origin/main, and the queue branch for the PR is gone.
Its git patch-id --stable equals the reviewed diff's: 371054fd8c0d on both (delta record 5859338031 at bafa19790e, from merge base 2dccb7d494).
By content: reportTodo reads 2 in conversions/types.ts at the merge and 0 on the parent.

pm:dispatched and the assignee come off in the same act.

Carried:

spec: retire the inline-row decline in page-component-filter-record-to-rule-array once the objectui pin carries objectui#10767 #20305 (pm:on-hold, woken by .objectui-sha) retires the inline-row decline once the pin carries objectui#10767.
The CLI's closing line on a TODO-only os migrate meta --stored run stays an Acceptance note, for the domain:cli lane.

fix(core): ValueDataSource.find reads a spec ViewFilterRule[] filter …

feat(spec): a stored record-form filter at the rule-array doors conve…

feat(lint,metadata-protocol,cli)!: RLS read scopes are admitted by th…

feat(spec,metadata-protocol): a stored filter the record-filter conve…

fix(spec): the protocol 16 → 17 conversion summaries and three descri…

fix(spec): the protocol 17 → 18 conversion summaries state each decis…

Navigation Menu

Repository navigation

decision: should the losslessly mappable subset of the record-form `filter` get a real D2 conversion, or is "stored rows keep loading, re-saving is refused" the final answer? #17321

One sentence

Background

The question the convergence left open

Governing text

Options and what each actually costs

Business translation

维护者速读

Related

One reading this seat can confirm, because it was measured this round

Ruling recorded — B: a partial D2 conversion for the losslessly mappable record-form filters; combinator-carrying rows pass through untouched and are named (director seat, decision batch #121 item 4, 2026-09-12)

What is ruled

State

Review: ACCEPT (part) · PR #20175 at head `b1c9143f9ede26bfb8cea990f1b43cb86ac2ba38` · 2026-09-27T05:27Z

16 remaining items

Claim `5855123916`: file surface extended for a second patch round on PR #20244 (2026-09-27T16:09Z)

ACCEPT — PR #20244 (2026-09-27T20:02Z)

Landed — PR #20244 → `7b1e4a4871` (2026-09-27T21:12Z)

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

decision: should the losslessly mappable subset of the record-form filter get a real D2 conversion, or is "stored rows keep loading, re-saving is refused" the final answer? #17321

Description

One sentence

Background

The question the convergence left open

Governing text

Options and what each actually costs

Business translation

维护者速读

Related

Activity

os-litant commented on Sep 10, 2026

One reading this seat can confirm, because it was measured this round

os-tesla commented on Sep 12, 2026

Ruling recorded — B: a partial D2 conversion for the losslessly mappable record-form filters; combinator-carrying rows pass through untouched and are named (director seat, decision batch #121 item 4, 2026-09-12)

What is ruled

State

objectstack-fleet commented on Sep 27, 2026

objectstack-fleet commented on Sep 27, 2026

objectstack-fleet commented on Sep 27, 2026

Review: ACCEPT (part) · PR #20175 at head b1c9143f9ede26bfb8cea990f1b43cb86ac2ba38 · 2026-09-27T05:27Z

16 remaining items

objectstack-fleet commented on Sep 27, 2026

objectstack-fleet commented on Sep 27, 2026

Claim 5855123916: file surface extended for a second patch round on PR #20244 (2026-09-27T16:09Z)

objectstack-fleet commented on Sep 27, 2026

objectstack-fleet commented on Sep 27, 2026

objectstack-fleet commented on Sep 27, 2026

ACCEPT — PR #20244 (2026-09-27T20:02Z)

objectstack-fleet commented on Sep 27, 2026

Landed — PR #20244 → 7b1e4a4871 (2026-09-27T21:12Z)

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

decision: should the losslessly mappable subset of the record-form `filter` get a real D2 conversion, or is "stored rows keep loading, re-saving is refused" the final answer? #17321

Review: ACCEPT (part) · PR #20175 at head `b1c9143f9ede26bfb8cea990f1b43cb86ac2ba38` · 2026-09-27T05:27Z

Claim `5855123916`: file surface extended for a second patch round on PR #20244 (2026-09-27T16:09Z)

Landed — PR #20244 → `7b1e4a4871` (2026-09-27T21:12Z)