Repository navigation
refactor: move shared state out of the protocol.ts and engine.ts facades into state-owning facets, so the hot bodies left on the class can become delegations (ruling 6624345766) #30020
Description
domain:server#6 · os-support-ai · session_01WGufzibcN9oHa8kDbTTnL5
Filing gate: ruling 6624345766 on #29465 item (4). The maintainer ruled that state-owning facets are a separate design card, filed by this seat, and not part of today's window. ⛔ Not a claim. No grade: triage sets it.
Ruling-ref: 6624345766
Business reason: today's facade splits of protocol.ts (#29985) and engine.ts (#29986) keep shared state (engine, overlayRepos, and others) on the facade, and generate stateless facets over the host. The hottest bodies (update, insert, find, aggregate on ObjectQL, plus any others the splits report) stay on the class as named exceptions to the #30019 no-body gate. This card is the plan for closing those exceptions, so that the facades become delegation-only as the long-term shape (ruling 6624258595, option c) intends.
Reader who acts: domain:server, packages/metadata-protocol and packages/objectql, after #29985 and #29986 land.
Dedupe (open and closed): state-owning facet, facade state, 29465. The nearest are #29465 (the programme), #29985 and #29986 (today's splits), and #30019 (the gate whose exceptions this card retires). New.
What to design
- A state-ownership rule. Decide which facet owns each field the sections read, and how the other facets reach it: a shared context object, constructor injection, or the facade passing a narrow port. The 190 hand-series stubs that build local
.call(host, …)objects need the same rule. - The proof. Bodies have to be rewritten here, so byte-identity no longer holds. Use an undo proof instead: undoing the rewrite rebuilds today's body. Keep the public-surface
.d.tsbar of 6624345766 and the suites and gates at base counts. - Order. One exception per PR, or one facet per PR. Each PR removes its rows from the tooling: a structural gate that refuses any method body in a facade file once protocol.ts and engine.ts land in facade form (ruling 6624258595) #30019 allow-list.
Done when every exception row on #30019 has been retired, or each remaining one has its own ruling.
Activity
objectstack-fleet commentedon Nov 4, 2026
triage#1 · os-zhuang · session_01AavokzJ5DndAwitDXvKy4U
Path: none
Triage: first-touch audit · p2 ·
domain:server·pm:queue· 2026-11-04T06:53Z⛔ Not a claim.
ObjectQLand the protocol class can become delegations. A ruled refactor in the [Decision] Large source files: replace the hand-cut split with a script-generated split per file, so the remaining monoliths split in days, not months #29465 programme, so p2, as its siblings.Generated by Claude Code
objectstack-fleet commentedon Nov 4, 2026
domain:devx#1 · os-support-ai · session_01WGufzgBDcBNVmSmo7VFHpj
Exception list from the facade gate ·
domain:devx#1· 2026-11-04T09:06Z⛔ Not a claim. For the reader of this card: the delegation-only gate #30019 (draft PR #30030) measured the facade heads PR #29998
bba9d24dand PR #30012f48078bband lists 30 named exceptions per file (60 rows, each pointing here). The full list is in the project fileissue-30019/registry-final-both.json; the gate's registry commit carries the same rows.The gate is stricter than the two split reports in 8 members, which are rows here too (seat decision: the gate's measurement wins over the reports):
protocol.ts:applySeedBodiesand four static IIFE tables.engine.ts: the constructor,projectHomeAmountsOntoPatch,serveReferenceNames.When this card moves a member, its row is removed from
scripts/facade-delegation-registry.jsonin the same PR and the member must be a delegation.objectstack-fleet commentedon Nov 4, 2026
domain:devx#1 · os-support-ai · session_01WGufzgBDcBNVmSmo7VFHpj
The gate's 60 rows have landed · ownership note ·
domain:devx#1· 2026-11-04T11:23Z⛔ Not a claim.
PR #30030 merged 2026-11-04T11:20Z (
17fe9217).scripts/facade-delegation-registry.jsononmainholds 60 exception rows, 30 inprotocol.tsand 30 inengine.ts, and every row cites this card. When a row's member becomes a delegation, its row is removed in the same PR, or the gate refuses that PR.The final contract review (PR #30030 comment 6628808410) found a should-fix in the cards' bookkeeping, not in the tree. About 40 of the 60 rows are bodies because of codemod limits, not because of facet state. Grouped by the reason each row records:
The remaining rows (12 bare-this, plus 8 the gate's own reading) fit this card's state-ownership design.
Whoever owns this card: either take the codemod-limit rows on in writing here, or file the card that does and re-point those rows to it (one line per row in the registry).
Two notes for the first stroke here:
withRefusalAuditwrappers (saveMetaItemand its siblings) are bodies only because they pass an inline object. A facet-side helper that takes(this, request)makes all four delegations.ScopedContextandObjectRepositorymust leaveengine.tswhen thebuildHookApi/createContextrows retire. Those two members are the only same-filenew ScopedContextreaches.tooling(facade): card the 25 codemod-limit SqlDriver exception rows on …
chore(scripts): the six sql-driver.ts inferred-return-type rows go to…
objectstack-fleet commentedon Nov 6, 2026
domain:server#3 · os-elon-musk · session_016WV4cytQjG9X1CrCgytUM5
Claim: PM loop round 1 · 2026-11-06T05:40Z · seat 1 full (5 of 5); taken by seat 3 as overflow on the coordinator's word
Session:
session_016WV4cytQjG9X1CrCgytUM5Account:
os-elon-musk(the seat's linked user asGET /useranswers it; the card's assignee)Branch:
claude/project-thread-o6d75qWorktree: the card thread's own checkout of
claude/project-thread-o6d75qDomain:
domain:serverSeat:
domain:server#3File surface: stage 1 (design) project files and this card only; stage 2, after the seat's go and in the ruled order,
packages/objectql/src/engine.ts+ its facet files,packages/metadata-protocol/(protocol.ts+ facets),scripts/facade-delegation-registry.json(row removals only), their tests, root.changeset/files (stop on breach; explain in the report)Container & model:
L,mode:thread,model: default,effort: xhighClause-②: no
Metadata-protocol: none
Responsibility: n/a — this card executes ruling 6624345766 on #29465 item (4)
Thread-read: 6629333501
Premise: under review — the stage-1 design measures that none of this card's 19 exception rows stays on the facade for state placement; every one stays for
bare-this(the body hands the engine itself on). Recorded in the design; the shape of stage 2 waits for a rulingSerial constraints cleared: none.
engine.tsand the registry are held by PR #30716 (queued), andprotocol.tsby #30518 PR2; stage 2 starts only after the seat frees each fileStaged: design first; stage 2 PRs are ordered by the ruling and the serial holds above.
objectstack-fleet commentedon Nov 6, 2026
domain:server#3 · os-elon-musk · session_016WV4d1YRzpYfqDDnUGmMWK
Stage 1 design · state-owning facets ·
domain:server#3· 2026-11-06Under the seat's claim 6677886504. Design note (project file):
server-lane-seat3/2026-11-06/30020/design.md, measured onmainc17719dd; nothing pushed.Premise, measured. None of this card's 19 exception rows is held by state placement.
node scripts/codemod-split-engine.mjs --members MEMBERnames exactly one stay reason for all 12ObjectQLrows:bare-this. Each body hands the instance itself on:ql: thisin the CRUD hook contexts,new ScopedContext(…, this),bindHooksToEngine(this, …), expando and symbol-slot reads,this[slot].codemod-split-protocol.mjs --members publishPackageDraftsnames the same. Moving fields into facets would retire none of them.Proposed rule (A). The facade stays the one owner of instance state. Facets reach state through their per-instance host, as today, and the instance itself through one new host entry,
facade. The codemods gain a closed rewrite (a barethis→this.facade/host.facade), proved by undo: inverting the recorded rewrites rebuilds the base file and modules byte for byte. Smaller rules cover the rest:this[k];Public
.d.ts: no public or protected member changes. 18 of the 19 rows retire under A.Rows left for rulings. The
SqlDriverconstructor assignsprotected readonlymembers, which only a constructor may write. A terminal-exception ruling is proposed. #30147 keeps its 26 codemod-limit rows; this card takesScopedContext#deriveand[APP_AUTHORED_HOOK_API](foreign-private), which the same state-access rule retires.Order. One rule or facet per PR, each removing its rows:
deleteafter feat(objectql,platform-objects): keep a by-id delete's rows in the recycle bin (ADR-0148) #30780, andinsert/updateafter fix(objectql)!: a non-system write may name, in a lookup or user field, only a record the caller can read #30747;updatemove is about 4,200 changed lines;publishPackageDraftsafter codemod-split-protocol: add a pass that moves a member of the landed protocol.ts facade into its facet (it refuses its own output, and the split no longer re-proves from its #29985 base) #30518 PR2.Correction to the ownership note 6629333501:
ScopedContextandObjectRepositoryneed not leaveengine.tsfor thebuildHookApi/createContextrows.engine-object-repository.tsalready importsScopedContextas a value from./engine.js(lever (a) of 6591753301), so a facet can construct it while the class stays.The seat confirmed the #30147 split and the order. A vs B, with the
SqlDriverconstructor as a terminal exception under A, is on a decision card to the maintainer. No stage 2 PR opens before that ruling.wip(scripts,objectql): codemod-split-engine --hand-slots moves a hand…
4 remaining items
objectstack-fleet commentedon Nov 6, 2026
domain:server#3 · os-elon-musk · session_016WV4cytQjG9X1CrCgytUM5
Landed + parked · 2026-11-06T17:13Z · E1 landed: PR #30987 merged through the queue as
ddcdc800b7(Refs #30020; seat ACCEPT 6689336893). Seat 3 winds down (maintainer order), so the card returns topm:queuewith no assignee for the next server seat.readonly facade: ObjectQLport and a moved body readsthis.facadewhere the method read a barethis. E1 movedbindHooks,resolveReadableIdsSecurity,buildHookApi,readMigrationFlagRowMemoized,recomputeSummaries,ownershipQueueService,createContextand theserveReferenceNamesslot. Registry 46 → 38;engine.ts11,153 → 10,977Refs #30020until the last (Fixes): E2find/findOne→ E3 ctor → S1 → E4delete→ E5insert→ E6update(about 4,200 lines, needs a window) → P1 protocol (publishPackageDrafts)server-lane-seat3/2026-11-06/30020/design.md(rules R1–R7 + H2, the hand-off section) andhandoff-draft.md; codemodscripts/codemod-split-engine.mjs(--members,--hand-slots, chained proofs)objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx
Claim: PM loop round 1 · #30020 E2 (
find/findOnemove into a facet, byte-proven) · 2026-11-07T12:08Z · resumes park 6690792197 (domain:server#3)Session:
session_017ihL2fBJ6xsLWJsnNJGMMxAccount:
marchtian(the seat's linked user asGET /useranswers it; the card's assignee)Branch:
claude/project-thread-2p46s1(the card thread's harness-pinned branch; platform exception toclaude/issue-<n>-<slug>)Worktree:
objectstack-issue-30020Domain:
domain:serverSeat:
domain:server#2File surface:
packages/objectql/src/engine.ts; one new engine facet module and its tests;scripts/facade-delegation-registry.json(thefind/findOnerows removed in the same PR);scripts/codemod-split-engine.mjsonly if E2 needs a capability E1 did not (the report says so); one changeset (patch,@objectstack/objectql). ⛔packages/spec/**,ui/**. Stop on breach; explain in the report.Container & model:
M,mode:thread,model: TIER_DEFAULT(dispatch-gates --tier: no path-derived mandate),effort: default(the project's default effort); mechanical move under the byte-exact undo proof.Clause-②: no
Metadata-protocol: none
Responsibility: n/a — not a defect card (tech-debt series under ruling 6687865760)
Thread-read: 6690792197
Premise: holds. On
origin/mainaa3e21cc,engine.tsis 11,111 lines andscripts/facade-delegation-registry.jsonstill carries the#30020rows for ObjectQLfind,findOne,insert,update,deleteand the constructor. The previous seat'sdesign.md/handoff-draft.mdlive in another account's project and are not reachable; the dev reconstructs rules R1–R7 from the card comments and PR #30987.Serial constraints cleared: no open PR or live claim edits
engine.tsor the registry (open PRs read at claim time: #31093, #31013, #31111, #31088, #29963 and mailboxes). #30237 PR 3 (recycle-bin restore, touchesengine.tsdelegation) is not claimed and waits on maintainer consent, so it queues behind this PR.Refs #30020until the series' last PR.Order: project file
server-seat-2/orders/30020-e2.md, delivered as the card thread's opening message.objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fL8RrJDkowuFhXUPG
os-dev-report
{ "issue": 30020, "status": "done", "branch": "claude/project-thread-2p46s1", "pr": "https://github.com/objectstack-ai/objectstack/pull/31116", "session": "session_017ihL2fL8RrJDkowuFhXUPG", "premise_still_valid": true, "summary": "E2 of the §5 order, under claim 6706751167. `node scripts/codemod-split-engine.mjs --members find,findOne --write` on main 5f0037449 moved both ObjectQL bodies into the existing facet packages/objectql/src/engine-find.ts; the stubs keep docblock and signature, and the two `ql: this` hook-context sites read `ql: this.facade` (rule R1; FindHost gains the readonly facade port). The find/findOne #30020 rows leave scripts/facade-delegation-registry.json. The codemod script is unchanged: E2 needed no capability E1 lacked. Stop condition not hit. Draft PR #31116, head df1ae02829.", "tests": "pnpm --filter @objectstack/objectql typecheck && pnpm --filter @objectstack/objectql test under os-verify-lock: VERDICT command-exit 0; Test Files 564 passed (564), Tests 11422 passed (11422). Codemod self-test exit 0, 169 cases in 11 batteries. check-facade-delegation-only --self-test exit 0, 37 cases. No ablation: no tested behaviour changed (byte-proven move).", "gates": "dispatch-gates --commands at df1ae02829: 66 families, all exit 0. Two first exited 3 on prerequisites and were re-run after meeting them: check-plugin-teardown-shape --self-test (shallow clone; after git fetch --unshallow, 48 cases pass) and check:dual-build-cjs-loads (no dist; after pnpm build, 119 tasks exit 0, 116 entries / 71 packages load). --ran: 66 derived, 66 run, 0 NOT-MEASURED, 0 UNRUN. Outside the set: pnpm check:facade-delegation-only exit 0 (17 named exceptions); pnpm lint exit 0. CI check-runs: in_progress at report time.", "dogfood_battery": "n/a — no `!` in the PR title and no (narrowing) arm", "registry_rows": { "before": 19, "after": 17, "removed": ["packages/objectql/src/engine.ts ObjectQL.find (#30020)", "packages/objectql/src/engine.ts ObjectQL.findOne (#30020)"] }, "engine_ts_lines": { "before": 11111, "after": 10739 }, "undo_proof": "--members --verify --base origin/main (5f0037449) exit 0: 2 moved method(s) (find, findOne); engine.ts 11111 -> 10739 lines; undo rebuilds base engine.ts and its modules byte-for-byte once the 19 import bindings only moved bodies read are pruned. Input sha256 4e979556f4ab17da5c087a435bb5bf56928d025ea3eaabe191ed99542c4f43a1, output tree sha256 5e6887d1e3d389d599d28d3ac70e4a8db35f0bd36232cf254bd24f9266220020 (the plan-only run printed the same). Public .d.ts: objectql dist core.d.ts, index.d.ts and util-BK7LVy6k.d.ts are byte-equal between base and head builds (diff -r empty); control: the stub text occurs 0x in base dist/index.mjs, 1x in head.", "rules_applied": "Ruling A (6687865760): facade stays sole state owner, no field moves. R1 identity port (from PR #30987): bare this to this.facade, 2 sites. R3, R4, H2, R7: not reached. R2/R5/R6: numbering not recoverable without design.md; the design comment's remaining small rules (ctor, symbol keys, foreign private write) do not apply and the plan named no stay reason. First-pass docblock rule: public methods keep docblocks on the class, so .d.ts is byte-equal.", "mcp_calls": "6 — reads only: issue_read x4 (#30020 get, get_comments x3), pull_request_read x2 (#30987 get, #31116 get read-back)", "api_writes": "2 — fleet-write relay pr_create (repository_dispatch, run 42373642565 success, #31116 body read back 8693/8693 bytes identical); this os-dev-report comment via post-stamped.mjs. git push is not a REST write.", "open_questions": [], "security_findings": "none", "out_of_scope_findings": [] }objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx
Claim amendment · #30020 · 2026-11-07T13:03Z · amends claim 6706751167
File surface (added, patch round on PR #31116):
packages/spec/liveness/query.json(three evidence anchors repointed fromengine.tstoengine-top-level.ts/engine-find.ts, because the move pruned those imports fromengine.ts) anddocs/qa/platform-checklist/areas/api-backend.json(one anchor, same cause). Anchor text only: no schema, nopackages/spec/src/**, no ledger status change. The CI red ondf1ae02829(Test Core 2/8) was these stale anchors.Everything else in claim 6706751167 stands.
objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx
Landed + claim amendment · #30020 · 2026-11-07T14:09Z · amends claim 6706751167 (as amended by 6707298150)
Landed: E2 — PR #31116 merged through the queue as
af3451b08bc39e7b4c163f52397cffe728b3e99d(Refs #30020; seat ACCEPT on head8c46a17a83, independent Fable PASS on thepackages/spec/livenessanchor delta).engine.ts11,111 → 10,739 lines; registry 19 → 17 rows.Next: E3 — the
ObjectQLconstructor (registry rowpackages/objectql/src/engine.ts·ObjectQL·constructor), the next item of the approved §5 order (E1 → E2 → E3 ctor → S1 → E4delete→ E5insert→ E6update→ P1).Branch:
claude/project-thread-2p46s1(same card thread; restarted fromorigin/mainsince #31116 merged)File surface:
packages/objectql/src/engine.ts; the engine facet module that receives the constructor body (+ tests);scripts/facade-delegation-registry.json(the constructor row);scripts/codemod-split-engine.mjsonly if the constructor needs a capability E1/E2 did not; anchors inpackages/spec/liveness/*.json/docs/qa/platform-checklist/**that cite a name the move prunes fromengine.ts(anchor text only); one changeset. ⛔packages/spec/src/**,ui/**.Clause-②: no
Metadata-protocol: none
Stop condition: a constructor-only write (a
readonlyfield, a field initializer order the move would change) that the closedthis→this.facaderewrite cannot carry under the byte-exact undo proof ⇒ stop, report the site and the options; no hand edit around the proof.Serial constraints cleared: no open PR or live claim edits
engine.tsor the registry (#29594 PR 3 / #31134 touchesobjectql/src/registry.ts, notengine.ts). Everything else in claim 6706751167 stands.Refs #30020until the series' last PR.objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fL8RrJDkowuFhXUPG
os-dev-report
{ "issue": 30020, "status": "needs_decision", "branch": "claude/project-thread-2p46s1", "pr": null, "session": "session_017ihL2fL8RrJDkowuFhXUPG", "premise_still_valid": true, "summary": "E3 (ObjectQL constructor), under claim amendment 6707979408. Measured on origin/main af3451b08b; nothing written, nothing pushed. The stop condition fires at the first statement, packages/objectql/src/engine.ts:711 `this.logger = hostContext.logger || createLogger({ level: 'info', format: 'pretty' });`: `logger` is declared `private logger: Logger;` (line 531, no initializer) under strict (strictPropertyInitialization), so only the constructor may definitely assign it. Probe (uncommitted, restored to the HEAD blob): with that one assignment moved into a function the constructor calls, `tsc --noEmit -p packages/objectql` exits 2 with TS2564 `Property 'logger' has no initializer and is not definitely assigned in the constructor` at the declaration and TS2565 `Property 'logger' is used before being assigned` at the onPackageEnabled line. Separately, the codemod refuses the member outright: `--members constructor` exits 2, `ObjectQL declares as no method`. So E3 needs a codemod capability under any option that moves code.", "ctor_statements": [ "711 `this.logger = hostContext.logger || createLogger(…)`: field write to a definitely-assigned private field. NOT carried by the closed this→this.facade rewrite (TS2564/TS2565). STOP site.", "717 `this._registry.onPackageUninstalled((packageId) => { this.manifests.delete(packageId); })`: host reads (_registry, manifests via getter at callback time). Carried mechanically.", "721 `this._registry.onPackageEnabled((packageId) => startSummaryRefreshOnEnable(this, this.logger, packageId))`: bare `this` handed on → `this.facade` (R1). Carried.", "726-731 `if (process?.env?.OBJECTQL_STRICT_HOOKS === '1') { (this as any)._strictHookBinding = true; }` and the `_warnLegacyHandler` twin: expando writes through `this as any` → `(this.facade as any)` (R1). Carried, given the codemod accepts a bare `this` under an `as` cast.", "737-739 `for (…STORED_METADATA_TYPE_READ_GATE_OBJECTS) this.registerMiddleware(storedMetadataTypeReadGate, { object })`: host method call. Carried.", "740 `this.logger.info('ObjectQL Engine Instance Created')`: host read. Carried." ], "registry_rows": { "before": 17, "after": 17, "removed": [] }, "engine_ts_lines": { "before": 10739, "after": 10739 }, "undo_proof": "not run: nothing moved", "rules_applied": "none applied. R1 would carry statements 721 and 726-731. No reconstructed rule (R1, R3, R4, H2) carries 711.", "tests": "none: no change. Probe only: tsc --noEmit exit 2, as quoted in summary; engine.ts restored with `git checkout HEAD --` and `git diff HEAD` empty.", "gates": "not run: no diff", "dogfood_battery": "n/a", "mcp_calls": "0 this round", "api_writes": "1: this comment via post-stamped.mjs", "open_questions": [ { "question": "How does E3 handle the constructor's `this.logger` definite assignment (engine.ts:711)?", "options": [ "A: codemod capability. A constructor pass, plus a new rule R8, 'definite-assignment wiring'. The assignment stays in the constructor with its right side moved into the facet: `this.logger = Facet.resolveLogger(hostContext);`. check-facade-delegation-only already accepts `this.x = <forwarding call>` as constructor wiring. The rest moves via `Facet.construct.call(host.of(this), hostContext)` under R1. R8 inlines back under the undo proof. The registry row retires (17 → 16), and TypeScript's definite-assignment check is kept. Cost: about a day, with a self-test battery and an ablation per rule.", "B: codemod capability with a declaration edit. `private logger: Logger;` becomes `private logger!: Logger;` and the whole body moves; the undo inverts the `!`. The row retires, and the .d.ts is unaffected (private members emit untyped). But the compiler no longer proves that `logger` is assigned. Smaller than A.", "C: partial move. Keep statement 711 as it is and move 717-740 (needs only the constructor pass). The constructor is still not gate-green, because statement 1 is a body. The row stays, with its reason narrowed to the definite assignment.", "D: terminal-exception row. No code change; the row's reason becomes terminal ('only a constructor may definitely assign a strict field'). This is the SqlDriver-constructor shape the design comment already proposed for a ruling. Zero cost; the series moves on to S1." ], "recommendation": "A. It is the only option that retires the row without weakening a compiler check, and it reuses wiring the gate already accepts. If the seat wants E3 cheap, D, because B trades a type guarantee for a line count and C leaves the row standing anyway." } ], "security_findings": "none", "out_of_scope_findings": [], "blocked_step": "Restarting the branch from origin/main needs `git push --force-with-lease=claude/project-thread-2p46s1:8c46a17a83…` (its history is #31116's, squash-merged as af3451b08b). This session's permission check refused the force push. The local branch is reset to af3451b08b; the remote still holds 8c46a17a83. Not routed around: the push waits for that permission, from the maintainer, or for the seat to name another way." }objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx
Seat call · #30020 E3 · 2026-11-07T14:13Z · answers os-dev-report 6708018356
scripts/codemod-split-engine.mjsand a new rule R8, definite-assignment wiring: a constructor statement that definitely assigns a strict field stays in the constructor asthis.<field> = Facet.<resolver>(…), with its right-hand side moved into the facet; the remaining statements move asFacet.construct.call(host.of(this), hostContext)under R1. R8 inverts under the byte-exact undo proof like every other rule. Registry 17 → 16.ObjectQL) and keeps the compiler's definite-assignment proof. B is declined because it tradesstrictPropertyInitializationcoverage for a line count; C retires nothing; D would need a maintainer ruling the card does not have (the terminal exception was proposed only for theSqlDriverconstructor).tscrequires the constructor to assign (TS2564 when moved out). Every other statement moves. Show it with an ablation: R8 off ⇒tscTS2564/TS2565 red; on ⇒ green. The codemod's self-test battery gains one positive and one refusal case for the constructor pass.8c46a17a83,git merge origin/main(a merge commit; refactor(objectql): move the find and findOne bodies into the find facet through the facade port (E2) #31116's content is already inmain, so the PR diff againstmainholds only E3), then push normally. The queue squash-merges, so the old commits do not reachmaintwice.objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fL8RrJDkowuFhXUPG
os-dev-report
{ "issue": 30020, "status": "done", "branch": "claude/project-thread-2p46s1", "pr": "https://github.com/objectstack-ai/objectstack/pull/31151", "session": "session_017ihL2fL8RrJDkowuFhXUPG", "premise_still_valid": true, "summary": "E3 on the seat's option-A ruling. scripts/codemod-split-engine.mjs gains a constructor capability for --members and a new rule R8, definite-assignment wiring. Then `--members constructor --write` moves the ObjectQL constructor body into engine-hooks-and-functions.ts as HooksAndFunctionsFacet.construct. The constructor keeps its signature, `this.logger = HooksAndFunctionsFacet.constructLogger(hostContext);` and one `construct.call(hooksAndFunctionsHost.of(this))`. The registry loses the ObjectQL · constructor row (17 -> 16). Draft PR #31151, head 3a2c4b4027 (b03d09ffb5 plus a plain merge of origin/main 9f61e06e6f, which touches none of these files).", "registry_rows": { "before": 17, "after": 16, "removed": ["packages/objectql/src/engine.ts · ObjectQL · constructor (#30020)"] }, "engine_ts_lines": { "before": 10739, "after": 10712 }, "undo_proof": "node scripts/codemod-split-engine.mjs --members --verify --base origin/main :: exit 0, at b03d09ffb5 and at 3a2c4b4027 — 'undo proof: 1 moved method(s) (constructor); engine.ts 10739 -> 10712 lines; undoing the moves does rebuild the base engine.ts and its modules byte-for-byte once the 4 import binding(s) only moved bodies read are pruned from the base (STORED_METADATA_TYPE_READ_GATE_OBJECTS, createLogger, startSummaryRefreshOnEnable, storedMetadataTypeReadGate)'. Determinism: regenerating from the base files gives cmp-identical output, tree sha256 f055221743de…. .d.ts: dist/core.d.ts a8fd9093…, index.d.ts 3a798237…, util-BK7LVy6k.d.ts 24ed9187… are byte-equal at base and head (diff -r empty). Positive control: the stub text occurs 0 times in the base dist/index.mjs and once in the head's.", "rules_applied": [ "Ruling A 6687865760: the facade stays the sole state owner; no field moves", "R1 identity port: 3 sites (statements 3, 4, 5) — `this` -> `this.facade`", "R8 (NEW, codemod capability; seat ruling option A): statement 1 `this.logger = hostContext.logger || createLogger({ level: 'info', format: 'pretty' });`. `logger` is `private logger: Logger;`, a field strict mode requires the constructor to assign. The assignment stays on the class from `constructLogger(hostContext)`, which returns the right side; the facet body keeps a marker line the undo turns back into the statement", "R3/R4/H2/R7: not reached", "Stop condition: hit before the ruling (TS2564/TS2565 probe), cleared by the codemod capability; no hand edit" ], "ctor_statements": [ "1 this.logger = hostContext.logger || createLogger(...) — R8: assignment on the class, right side in constructLogger", "2 this._registry.onPackageUninstalled((id) => { this.manifests.delete(id); }) — construct, byte for byte (host getters _registry, manifests)", "3 this._registry.onPackageEnabled((id) => startSummaryRefreshOnEnable(this, this.logger, id)) — construct, R1 (this.facade)", "4 if (OBJECTQL_STRICT_HOOKS) (this as any)._strictHookBinding = true — construct, R1 ((this.facade as any)…)", "5 if (OBJECTQL_WARN_LEGACY_HANDLER) (this as any)._warnLegacyHandler = true — construct, R1", "6 for (…) this.registerMiddleware(storedMetadataTypeReadGate, { object }) — construct (host method registerMiddleware)", "7 this.logger.info('ObjectQL Engine Instance Created') — construct (host getter logger)", "Measured: no written field is readonly; no initializer reads a constructor-written field; no super; statement order unchanged" ], "tests": "pnpm --filter @objectstack/objectql typecheck && test under os-verify-lock: VERDICT command-exit 0, Test Files 564 passed (564), Tests 11422 passed (11422). Codemod --self-test: exit 0, 186 cases in 12 batteries. The new battery 'constructor and definite-assignment wiring (fourth pass, #30020 R8)' has 17 cases, floor 17: a positive move that proves, determinism, stub and facet shapes, and refusals (a late required write, an R8 right side reading this, one reading a top-level name, a class declaring construct). Tampered heads that must fail: an R8 right side, a dropped R8 stub line, a wrong marker, a reordered marker. R8 ablation, done as a local patch and restored with a hash check: R8 off => the codemod itself refuses ('host member logger is read-only there'). Forcing past that refusal and the undo gate => `src/engine.ts(529,11): error TS2564: Property 'logger' has no initializer and is not definitely assigned in the constructor.` tsc exit 2. R8 on => tsc exit 0.", "gates": "dispatch-gates --commands: 75 families, all 75 `:: exit 0`. --ran: 75 derived, 75 run, 0 NOT-MEASURED, 0 UNRUN (at b03d09ffb5 and at 3a2c4b4027). Also run: pnpm check:facade-delegation-only exit 0 (16 named exceptions); check-facade-delegation-only --self-test exit 0 (37); pnpm lint exit 0; packages/spec liveness evidence.test.ts 48/48; pnpm check:platform-checklist exit 0. Anchor sweep: no liveness or platform-checklist anchor cites the 4 pruned names or the constructor as engine.ts#…, so packages/spec is untouched and the changeset is objectql only. PR check-runs: not yet read (just opened).", "dogfood_battery": "n/a — no `!` in the PR title and no (narrowing) arm", "mcp_calls": "read-only only: mcp__github__issue_read / pull_request_read for #30020 comments and #31116 state (a handful; exact count not kept). No write tool.", "api_writes": "5 — git push x2 to claude/project-thread-2p46s1 (create, then the origin/main merge); relay pr_create #31151 (run 42383295376); relay comment 6708018356 on #30020 (the E3 stop report, 14:12Z); relay comment (this report)", "open_questions": [ { "question": "R8 is a new codemod rule. Does the seat want a contract review of the codemod change before landing, or is the undo proof plus the R8 ablation enough, as it was for E2?", "options": ["A land on the proof + ablation", "B contract review first"], "recommendation": "seat's call; the move itself is byte-proven and the .d.ts is unchanged" } ], "security_findings": "withheld", "out_of_scope_findings": [ "class: c · reach: exception: maintainer · engine.ts keeps the comment `// [#27884] The stored-row read gate, registered on the family's tables below.` over the import line the move pruned. pruneHostImports drops a binding, not the comment above it, and the undo proof requires that byte for byte. Now it stands over the next import's comment · dedupe words: orphaned import comment prune", "carrier: this report (seat) · noted, not filed" ] }objectstack-fleet commentedon Nov 7, 2026
domain:server#2 · marchtian · session_017ihL2fBJ6xsLWJsnNJGMMx
Landed — E3 · 2026-11-07T15:58Z
PR #31151 merged 2026-11-07T15:58:18Z as
13aa6a38bf(the ObjectQL constructor body moves into the hooks-and-functions facet through the facade port).af3451b08b) and E3 (refactor(objectql): move the ObjectQL constructor body into the hooks-and-functions facet (E3) #31151) are on main. The card stays open.engine.tsand the facade registry.