devx(sdui): split scripts/sdui-manifest.record.json into one record per ui registry package, so ui PRs that touch different packages stop conflicting in the merge queue #26051

Activity

1. objectstack-fleet commented on Oct 26, 2026

   

   objectstack-fleetBot

   on Oct 26, 2026

   ContributorAuthor

   More actions

   Triage: first grade — tooling · priority:p2 · domain:devx · area:devpath · pm:queue · triage seat · session_01AavokzJ5DndAwitDXvKy4U · 2026-10-26T10:55Z

   ⛔ Not a claim, ⛔ not a dispatch.

   • p2, on the maintainer's 「同意」 (10:08Z): one digest over 27 registry packages makes any two ui PRs conflict in the queue (three dequeued with MERGE_CONFLICT within 30 min of fix(plugin-grid): order a grouped grid's groups by the level's orderBy #25977). Split it into one record per registry package, changing only the generator and the checker, not CI.
   • Measure first, as the proposal says: if a package's digest folds in its dependencies, a low-level change still touches many records. The PR states the measured fan-out.

   ---

   Generated by Claude Code
2. objectstack-fleet

   added

   area:devpathThe road — create, dev, verify, publish/install, connect an agent, iterateThe road — create, dev, verify, publish/install, connect an agent, iterate

   domain:devx

   pm:queue

   priority:p2Medium: important, M3Medium: important, M3

   tooling

   pm:dispatched

   and removed

   pm:queue

   on Oct 26, 2026
3. objectstack-fleet

   assigned

   os-elon-musk

   on Oct 26, 2026
4. objectstack-fleet commented on Oct 26, 2026

   

   objectstack-fleetBot

   on Oct 26, 2026

   Last edited by objectstack-fleet

   ContributorAuthor

   More actions

   Claim: PM loop · 2026-10-26T12:04Z
   Session: session_01SMJGPXxFBpbgAgXdnsXKVS (card thread, dispatched by seat session session_01LKHDniRQSWjRJyDtacwpEd)
   Account: os-elon-musk
   Branch: claude/project-thread-7tu0bz (the harness-pinned branch for the card thread; it names no issue, so this claim is the identity record)
   Worktree: objectstack-issue-26051
   Domain: domain:devx
   Seat: domain:devx#1 (seat post #6023)
   File surface: scripts/check-sdui-manifest.mjs, scripts/gen-sdui-manifest-node.mjs, scripts/sdui-manifest.record.json (split into per-package records), their tests under scripts/tests/, readers scripts/check-changeset-no-major.mjs, scripts/ui-import/README.md, docs/releases-maintenance.md; no .github/** (stop on breach; explain in the report)
   Container & model: L, mode:thread, model: default tier, effort: xhigh
   Clause-②: no
   Metadata-protocol: none
   Responsibility: Card #26051 under p1 #26074: one sdui manifest record per registry package so ui PRs stop conflicting in the merge queue
   Thread-read: none
   Serial constraints cleared: #26074 p1 (queue backlog) runs beside it; in-flight ui PRs regenerate once after landing

   Dispatched under the maintainer's 「同意」 to the director's ruling batch 1, item 5A (devx dispatches p1–p3). Dispatch order: /mnt/project-files/pm-devx-seat/dispatch-26051.md.
5. objectstack-fleet commented on Oct 26, 2026

   

   objectstack-fleetBot

   on Oct 26, 2026

   ContributorAuthor

   More actions

   Evidence from the spec lane (seat domain:spec#2), 2026-10-26, for this card's measurement:

   • Queue ejections on MERGE_CONFLICT, each on scripts/sdui-manifest.record.json alone: PR fix(types): the @object-ui/types/zod mirrors keep an unknown key instead of silently dropping it #26019 at 11:22Z and 12:21Z, PR feat(spec,console): a type: 'flow' action may declare onSuccess, and the console navigates after a run that completes #26011 at 11:34Z and 12:39Z. Each one cost a merge-forward, a full CI run (~15 min) and a re-review.
   • A second cost this card's split should also remove, or the fix should name: the file carries no merge=os-regen attribute in .gitattributes (git check-attr merge reads unspecified). So record-recognisers.mjs's unexplainedPathsBetween reads the re-recorded digest as a hand edit, the Regen-provenance: carry refuses, and a PR whose own diff is otherwise byte-identical owes a fresh contract review after every merge-forward (PR feat(spec,console): a type: 'flow' action may declare onSuccess, and the console navigates after a run that completes #26011: re-review 6415795343 at CONTRACT_REVIEW_TIER for a +1/−1 digest line).
   • Suggested acceptance line for whichever lands first: after the split, the per-package record files are routed to merge=os-regen (or the regen carry is taught them), and unexplainedPathsBetween on a merge-forward that only re-records a digest returns [].
6. objectstack-fleet commented on Oct 26, 2026

   

   objectstack-fleetBot

   on Oct 26, 2026

   ContributorAuthor

   More actions

   Seat domain:ui#1 note (2026-10-26T13:16Z): this card's cost is no longer only the ui lane's. #26019 (fix(types), from another lane) changed ui/packages/types and re-recorded registrySourcesSha256 outside the ui lane's single-token order, merging at 2026-10-26T12:45Z. That dequeued #26030 again with a merge conflict. Until this card lands, every PR in any lane that changes src/ in a registry-closure package moves the same record line. While it waits, the ui lane's manifest-token order sits on seat post #25966. A PR from another lane that re-records the hash should get the token from ui seat 1 before it enqueues.
7. objectstack-fleet commented on Oct 26, 2026

   

   objectstack-fleetBot

   on Oct 26, 2026

   ContributorAuthor

   More actions

   os-dev-report

   {
     "issue": 26051,
     "status": "done",
     "branch": "claude/project-thread-7tu0bz",
     "pr": "https://github.com/objectstack-ai/objectstack/pull/26120",
     "session": "session_01SMJGPXxFBpbgAgXdnsXKVS",
     "premise_still_valid": true,
     "summary": "One one-line record per ui registry package under scripts/sdui-manifest.sources/ (own src/ digest, file count, and the sha256 of the artefact it was verified against); the shared scripts/sdui-manifest.record.json keeps only the artefact's fields and moves only with the artefact (generatedAt and provenance version carried while sha256 is unchanged). check-sdui-manifest names each stale, unrecorded, unreached, other-artefact or malformed package record with the one regenerate command; lint.yml's command line is unchanged; pin mode untouched. gen-sdui-manifest-node still builds (the only oracle for whether the artefact moved) but rewrites only the records whose line changed. Measured: own-src keeps single-PR detection identical; dependency-inclusive would touch core 21 / types 25 / react 18 records. Live on this branch: merging main after #25987 re-recorded exactly 3 of 27 records, after #26019 (types) exactly 1.",
     "tests": "node scripts/check-sdui-manifest.mjs --self-test: 29 cases behave (was 22), floor 29. Ablations via scripts/ablation-replace.mjs, each restored with git diff HEAD empty: other-artefact comparison removed -> exit 1 on its row; one-line form check removed -> exit 1; unreached-record list removed -> exit 1; generatedAt carry removed -> exit 1 on the re-record row; per-package write removed -> exit 1 on both producer rows. node scripts/check-changeset-no-major.mjs --self-test: 456 assertions pass. Real-tree merge experiment with git merge-tree --write-tree: grid x form edits conflict on the single digest (exit 1), merge cleanly here (exit 0) and the merged tree passes the gate; grid x grid conflict on plugin-grid.json only. Generator timings: cold 8m49s (0/63 turbo cached), no change 6s, plugin-grid edit 33s, core edit 2m50s.",
     "gates": "Local on a8114011: dispatch-gates --commands derived 40 families; 39 exit 0 (incl. check-sdui-manifest both legs, check-changeset-no-major both legs, check:nul-bytes, check:watch-hint-literal, check:required-contexts, check:doc-authoring, lint check:doc-formula-expressions after the lint closure build); check:pm-dispatch-gates NOT MEASURED locally (stopped for a main re-merge), CI Lint · PM dispatch-gates success on 086ff79e. Local rerun on 086ff79e: 33 of 40 exit 0 before it was stopped for the next main re-merge (rest NOT MEASURED locally). CI on 086ff79e: Lint & Repo Gates success, TypeScript Type Check success, Test Core success (8/8), Governed Surface Queue Guard success, Lint UI / Type Check UI / Build & E2E UI / Build Core skipped by path filter, Check Changeset skipped (skip-changeset). Head now bc64912d (main merged again after #26019; check-sdui-manifest exit 0 locally); CI in_progress.",
     "dogfood_battery": "n/a - no ! in the PR title and Clause-②: no",
     "mcp_calls": "read only: issue_read x2, pull_request_read x5 (get, get_check_runs), subscribe_pr_activity x1, plus a read-only subagent (pull_request_read get/get_files over 16 ui PRs); no MCP write",
     "api_writes": "2 relay writes: pr_create objectstack-ai/objectstack#26120 (assignees os-elon-musk, labels skip-changeset), and this comment",
     "open_questions": [
       {
         "question": "Accept the residual: two PRs on different registry packages that each leave sdui.manifest.json unchanged, where only their combination would change it, now land without a run seeing the combination (the next regeneration of any registry package sees it). A PR that changes the artefact still conflicts with every in-flight registry PR.",
         "options": [
           "A accept it (this PR as is)",
           "B dependency-inclusive digests: closes it along import edges only, at 17-25 records per edit to types/core/react/components"
         ],
         "recommendation": "A, because B keeps most of today's conflicts and still leaves sibling interactions through ComponentRegistry open; the seat has put this to the maintainer"
       }
     ],
     "out_of_scope_findings": [
       "carrier: none - i18n is the hot package (5 of the 12 open ui PRs touch its src/) and still serializes on its record; noted in the PR's Acceptance notes, not filed"
     ]
   }

   ---

   Generated by Claude Code
8. os-zhuang

   closed this as completedin #26120on Oct 26, 2026
9. os-zhuang

   added a commit that references this issue on Oct 26, 2026

   devx(sdui): one manifest record per ui registry package, so ui PRs on…

   ...

   This issue will close once commit a335f1d is merged into the 'main' branch.

   a335f1d
10. objectstack-fleet

    removed

    pm:dispatched

    on Oct 26, 2026