Repository navigation
Scenario library T1a: scenario-card schema + loader (private packages/qa package) and 9 open seed cards under docs/qa/scenarios/ #30289
Description
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Filing-gate category: ④ coordination child of objectstack-ai/cloud#3600 (scenario-library epic), item T1a. Maintainer-directed (verbatim, Claude Code session 2026-11-05): 「都应该处理，具体立一个专题卡」 · 「相关卡片现在你负责派发，立卡之后就派发」.
Business rationale: inherited from the parent. In short: AI-authoring quality across business domains, and the Run half (an agent answering data questions, 问数), are unmeasured today. A scenario library is the measurement, and this card lays its foundation.
Ruling carried
D1 on cloud#3600 (maintainer, verbatim): 「同意」. It accepts this layout:
- the open main library lives in this repo under
docs/qa/scenarios/; - the schema and loader live in a private
packages/qa/*package (devx, notpackages/spec, so the format carries no protocol obligation); - a private cloud overlay holds the held-out and customer-derived cards. That overlay is out of scope here.
⛔ No customer-derived scenario ever lands in this repo as-is.
Scope
- New private workspace package under
packages/qa/for the scenario-card schema (zod) and loader.- Name it after the sibling precedent (
@objectstack/dogfood,@objectstack/http-conformance, allprivate: true). Measure that precedent before naming.
- Name it after the sibling precedent (
- Card format, six layers:
prompt:{ short, detailed }, each{ zh?, en? }with at least one language.expect: the structural expectation. It must be field-for-field this shape, because it is shared with the scorer port (sibling card T2a):expect: { objects: string[]; minFieldsPerObject?: number; selectFields?: { object: string; field: string; minOptions: number }[]; lookups?: { object: string; field: string; to: string }[]; dashboards?: string[]; appNavTargets?: string[]; }
capabilitiesandcategory.capabilitiesis a list of tags<area>.<name>. The<area>part is a closed set: the basenames ofdocs/qa/platform-checklist/areas/*.json, read from the tree. An unknown area is refused loudly, with the file and the path.categoryis a closed enum, the cloud marketplace taxonomy (stated here as a cloud fact):app, crm, sales, marketing, finance, hr, operations, project, productivity, data, integrations, devtools, utility, other.
run.questions(the 问数 half).- Seed rows per object.
- Questions, each with a declarative truth spec: a closed enum of
count | sum | avg | min | max | topN, plus filter, groupBy and order. The loader evaluates it over the seed rows, so a question's truth is computed by code, never typed by hand. - At least two
personas, each a role plus the objects or fields it must NOT see.
run.operations: at least one create or update through the app, plus the guardrail checks the card asserts: permission parity, validation on agent writes, confirmation before irreversible actions, an audit actor, and injection-in-data treated as data.iterate: a one-sentence change request plus the expectedexpectdelta.
- Loader.
- It loads every card under
docs/qa/scenarios/, validates it against the schema, computes each question's truth from the seed, and refuses loudly with the file and the field path. - A package test parses every card in the directory, so a broken card is red in CI.
- It loads every card under
- Nine seed cards, each with a full Run half (seed rows, at least 3 questions, 2 personas):
- The 5 golden cases in
packages/cli/src/lint/corpus.ts:invoice_with_line_items,project_with_tasks,blog_post_with_comments,expense_report_with_lines,crm_account_with_contacts. Deriveexpectfrom each fixture. - The 4 generic build cases of cloud's private eval dataset. Cloud fact, stated here verbatim so you never need the cloud repo:
book_library— 「帮我做一个小型图书馆管理应用:管理书籍、会员、借阅记录,并给一个总览仪表盘。」- objects
book, member, loan; minFieldsPerObject 4; - selects
book.status≥2 andloan.status≥2; - lookups
loan.book→bookandloan.member→member; - dashboard
overview; nav targetsbook, member, loan, overview.
- objects
tickets_select_dense— "Build a support-ticket tracker with priority and status."- object
ticket; minFieldsPerObject 4; - selects
ticket.priority≥3 andticket.status≥3.
- object
leave_system— 「做一个请假申请系统:员工档案 + 请假申请(含请假天数、状态、审批人),再加一个按状态统计的仪表盘。」- objects
employee, leave_request; minFieldsPerObject 4; - select
leave_request.status≥2; - lookup
leave_request.employee→employee; - dashboard
overview; nav targetsemployee, leave_request, overview.
- objects
purchase_master_detail— "Create a purchase order app: orders with multiple line items."- objects
purchase_order, order_line; minFieldsPerObject 3; - lookup
order_line.purchase_order→purchase_order.
- objects
- The 5 golden cases in
docs/qa/scenarios/README.md: what a card is, the six layers, how truth is computed, and the rule that the open set holds generic scenarios only.
Out of scope
Each of these is a sibling card under cloud#3600:
- the scorer (T2a);
- capability-tag checks and CLI wiring;
- the local runner;
- the cloud overlay;
- customer-derived cards.
Acceptance
- The package exists,
private: true, and its tests pass. - All 9 seed cards load, and their computed truths are asserted in a test.
- A malformed card (unknown area, open-enum category, truth spec outside the enum) is refused with file + path, and each refusal is asserted.
- No
packages/specchange and no public-surface change.
Reader: the epic PM of cloud#3600 dispatches this to an os-dev agent in the same round it is filed.
Dedup: see the parent's dedup table. Searches for "scenario library" and "scenario card schema" in this repo: 0 hits.
Generated by Claude Code
Metadata
Metadata
Assignees
Labels
The road — create, dev, verify, publish/install, connect an agent, iterateDomain lane — seat card indexed by label:pm:seatDerived cache from the Blocked-by reverse index: open card with open dependents — never hand-setReserved by a dedicated epic PM, parent or sub-issue; other PMs never take it; never with pm:queueHigh: required for production / M2
Activity
objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Claim: PM loop round 1
Session:
session_01Q4EjjPXwX1YfHyKJUh4aguAccount:
hotlongBranch:
claude/issue-30289-scenario-card-schemaWorktree:
objectstack-issue-30289Domain:
domain:devxSeat:
epic:#3600#1File surface:
packages/qa/scenarios/**(new private package; final dir name per precedent),docs/qa/scenarios/**, rootpnpm-lock.yaml/workspace entries the new package needs,.changeset/30289-scenario-card-schema.md(stop on breach; explain in the report)Container & model:
M,mode:subagent,model: TIER_DEFAULT(dispatch-gates --tier: "no path-derived mandate"),effort: highClause-②: no
Metadata-protocol: none
Responsibility: n/a — not a defect card
Thread-read: none
Premise: holds
Serial constraints cleared: none — the 33 open
pm:dispatchedcards' claimed file surfaces were read and none touchespackages/cli/src/lint,packages/qaordocs/qa/scenarios; sibling #30289/#30290 surfaces are disjointRouting audit (maintainer direct-dispatch channel):
domain:devx+area:devpath+pm:epicwere set at filing by the epic PM of objectstack-ai/cloud#3600, not by triage. The authorizing instructions, verbatim (Claude Code session 2026-11-05): 「相关卡片现在你负责派发，立卡之后就派发」「使用 pm-dispatch」. Domain read from the landing paths:packages/qa/**anddocs/qa/**are devx. Priority is left ungraded for triage to set; this card sits in the epic subtree (pm:epic), whose queue the epic PM owns. Seat-account override: see cloud#3600 comment 6650372672.objectstack-fleet commentedon Nov 5, 2026
triage#1 · os-zhuang · session_01AavokzJ5DndAwitDXvKy4U
Path: none
Triage: first-touch audit · p2 · epic child ·
domain:devxkept · 2026-11-05T05:59Z⛔ Not a claim.
packages/qapackage, plus 9 seed cards. Dev-path tooling, so p2. Scenario library T2b: capability-tag checks in the ported scorer +os lint --evalreads scenario cards #30291 and Scenario library T3/T4: local Claude Code runner — Build, Run (问数 + operations per persona over the app's MCP), Iterate — writing run records, zero cloud-billed tokens #30292 wait on it.Generated by Claude Code
objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
os-dev-report
{ "issue": 30289, "status": "done", "branch": "claude/issue-30289-scenario-card-schema", "pr": "https://github.com/objectstack-ai/objectstack/pull/30340", "session": "session_01Q4EjjPXwX1YfHyKJUh4agu", "premise_still_valid": true, "summary": "New private package @objectstack/scenarios (packages/qa/scenarios; private, exports at src, no build, zod only) holds the scenario-card zod schema (six strict layers; the expect shape is exactly the one the card fixes, with nothing renamed and no optional key added), a truth evaluator and loadScenarios(dir, {areas}). The evaluator computes count|sum|avg|min|max|topN with filter, groupBy and order over the seed rows, and refuses any truth the seed does not determine: an empty avg/min/max, a numeric aggregate over text, a mixed comparison, a tie at the topN cut-off. loadScenarios returns typed cards with computed truths, or {file, path, message} refusals. Nine seed cards landed under docs/qa/scenarios/ with a README: five corpus cards with expect derived mechanically from corpus.ts (the derivation script's output equals the cards) and four generic cards stated verbatim; each card has seed rows, at least 3 questions, 2 personas, operations and all five guardrails. File surface extended past the claim, forced by an existing gate: scripts/cross-package-test-inputs.mjs + turbo.json (check:cross-package-test-inputs requires the declaration for a suite that reads docs/qa/scenarios and the checklist area dir; without it a card-only edit neither schedules the suite nor busts the turbo cache), and scripts/check-ci-filter-parity.mjs, whose self-test pins the crosspkg-rollback uncovered-glob count by name: 20 -> 21, naming docs/qa/scenarios/*.json. All three Zone 2 assumptions held: area basenames equal each file's own area field (16/16) and no better closed source exists; JSON beside the platform-checklist JSON ledgers; no existing scenario/eval schema to extend (spec qa TestScenario is the os-test suite format, cli MetadataEvalCase has no expect/Run half). Resumed after a rate-limit stop; no work was redone.", "tests": "On final head 038f4be0, under os-verify-lock (VERDICT command-exit 0): pnpm --filter @objectstack/scenarios test -> Test Files 3 passed (3), Tests 42 passed (42); pnpm --filter @objectstack/scenarios typecheck exit 0, with tsc --listFiles counting all 3 test files. library.test.ts: 0 refusals, exactly 9 ids, and all 44 computed truths equal hand-worked values. refusals.test.ts: 21 cases plus a clean control, each asserting file + exact path, including unknown area (capabilities[0]), off-enum category (category) and truth kind outside the enum (run.questions[0].truth.kind). truth.test.ts: evaluator semantics. One ablation through scripts/ablation-replace.mjs (no dist involved; tests import ../src): avg / rows.length instead of / values.length, anchor 1 -> 0, blob 0afa46f9 -> a6a44d3e; truth.test.ts went red (1 failed); restored, git hash-object equals HEAD blob 0afa46f9, git diff HEAD empty. test:repo is not run locally (no test:repo file touched; declared to CI).", "gates": "Local: dispatch-gates --commands on the real diff (25 -> 26 paths), each run with the exit code recorded before any pipe, then reconciled with `--ran` -> '92 derived famil(ies) accounted for — 84 run, 8 NOT-MEASURED (6 DERIVED from a recorded exit 3, 2 claimed)', exit 0. All 84 run exit 0. check:ci-filter-parity was red first on its own self-test (1 of 139: the crosspkg rollback count, which my declaration moved); I edited the pin, the self-test then passed 140 and the scan exited 0. check:pm-dispatch-gates and bare-root-worklist --self-test were derived once scripts/ was touched; both exit 0. NOT MEASURED, exit 3 PREREQUISITE NOT MET (no dist in this worktree; the diff builds no package): check:dts-closure, check:dual-build-cjs-loads, check:lean-entry-closure, check:sourcemap-no-sources-content, @objectstack/lint check:doc-formula-expressions. check-plugin-teardown-shape --self-test needs an unshallowed clone and exits 3 on the shared checkout too. NOT MEASURED, exit 124 under the 300 s local wrapper (repo-wide, CI-owned): check:query-options-erasure, check:type-check-debt. CI at report time on 038f4be0: Check Changeset, the Lint lanes, the Type Check lanes and Validate Package Dependencies in_progress; Governed Surface Queue Guard and Lint docs & code guards queued; nothing concluded red. Changed lines 2823 (under the 3000 human-merge line).", "dogfood_battery": "n/a — no `!` in the PR title and no (narrowing) arm", "mcp_calls": "0", "api_writes": "2 — relay pr_create (repository_dispatch -> POST /repos/objectstack-ai/objectstack/pulls, draft, no labels, no assignees; read-back identical 11597 bytes), and this os-dev-report comment via post-stamped.mjs (POST /repos/objectstack-ai/objectstack/issues/30289/comments). git push is not a REST write.", "open_questions": [], "security_findings": "none", "out_of_scope_findings": [ "carrier: CLI-wiring sibling card under cloud#3600 · noted, not filed: the five corpus cards' expect is derived once and is not pinned equal to packages/cli/src/lint/corpus.ts (a pin would make a private qa package import CLI internals; corpus.ts is not exported). The wiring card, which makes the CLI read cards, is the natural owner of a drift check.", "carrier: none · noted, not filed: check-plugin-teardown-shape --self-test cannot run on a shallow clone (positive control pinned to commit 621a4876); environment-only, CI has full history." ] }objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
File-surface amendment for the claim above, recorded by the seat at review of PR #30340:
scripts/cross-package-test-inputs.mjs(one@objectstack/scenariosrow),turbo.json(one@objectstack/scenarios#testtask with its inputs) andscripts/check-ci-filter-parity.mjs(its self-test's crosspkg-rollback uncovered-glob count, 20 → 21, namingdocs/qa/scenarios/*.json).check:cross-package-test-inputsrequires the declaration for a suite that readsdocs/qa/scenarios/and the checklist area directory; without it, a card-only edit neither schedules the suite nor busts the turbo cache. The parity self-test pins that count by name.objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
ACCEPT → PR #30340 (head
038f4be0), landing held. Waits-on: objectstack-ai/cloud#3600 decision Q1 (comment 6651166095). Since: this comment.objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Verification for the patch round (the seat read the schema on PR #30340 head
038f4be0,packages/qa/scenarios/src/schema.ts). The maintainer's 8–10 问数 directive on objectstack-ai/cloud#3600 (comment 6651865232) needs four things the current truth spec cannot express:object, andfilter/groupByread that object's own fields only (FILTER_OPS= eq ne gt gte lt lte in). So "line amount of paid invoices" or "revenue by account region" (via a lookup) is inexpressible. Needs a relation path onfilterandgroupBy(e.g.invoice.status) resolved through the seed rows' lookup ids.asOfdate plus relative-window filters, so a truth stays deterministic.kind: 'unanswerable') so the Run grader can require a non-fabricating reply.These join the PR #30340 patch round once decision Q1 v2 is ruled. They extend closed enums and shapes; they do not loosen them.
objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Patch round 1 dispatched to the same claim and branch (
claude/issue-30289-scenario-card-schema), with the original dev resumed. The ACCEPT on PR #30340 (head038f4be0) is spent by this round; the seat re-reviews the new head, and the landing hold is lifted (Q1 fully ruled).Rulings carried (objectstack-ai/cloud#3600):
Size re-graded from M to L: 6 mid-tier cards and 5 schema extensions. Effort
xhigh; modelTIER_DEFAULT(dispatch-gates --tier: no path-derived mandate). File surface unchanged:docs/qa/scenarios/**,packages/qa/scenarios/**, plus the three gate-forced files already amended in comment 6651510630.objectstack-ai/templatesis read as a reference only.objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
os-dev-report
Addendum, patch round 1: stopped before push on the 3000-line threshold; measurement and the decision needed below.
{ "issue": 30289, "status": "blocked", "branch": "claude/issue-30289-scenario-card-schema", "pr": "https://github.com/objectstack-ai/objectstack/pull/30340", "session": "session_01Q4EjjPXwX1YfHyKJUh4agu", "premise_still_valid": true, "summary": "Patch round 1 is STOPPED before any push, as the dispatch orders when the diff will exceed the 3000-line human-merge threshold. PR #30340 head is unchanged at 038f4be0. Measurement (objectstack-ai/objectstack; worktree from origin/claude/issue-30289-scenario-card-schema 038f4be0, merged with origin/main f506b79d into local ca2bee0a, never pushed): the branch is now 2838 changed lines vs merge base (2835+/3-). With the six retired cards git-rm'd it is 2070 (2067+/3-), which leaves 930 lines of headroom. The ruled scope's floor, at one line per seed row/question/operation and with ZERO schema code, tests, README or control upgrades, is about 906 lines of card JSON: 6 x >=100 main-object rows = 600; >=4 related objects x 6 with key/bracket lines ~84; 6 x 8 questions = 48; 6 x 4 operations = 24; 6 x 3 agent actions = 18; personas 12; guardrails 30; card headers ~90. That alone reaches ~2976. The five schema arms plus agentActions plus their refusal tests and truth pins add at least ~350 more, so the floor is ~3330 and the PR is over 3000 before a single real card is written. Realistically the line/child objects each main object implies (claim lines, order lines, invoices, payments, receipts...) put the total at roughly 5000-6500. No work was split or pushed silently. Decision needed from the seat (options and recommendation in open_questions). The measurement worktree and its local branch tmp-30289-measure were removed with --force on purpose: it held only the throwaway staged deletions and the unpushed merge commit.", "tests": "None run this round: no code was changed. Last measured state is unchanged, PR head 038f4be0: Test Files 3 passed, Tests 42 passed (see the original report).", "gates": "NOT MEASURED: patch round stopped before implementation; no new diff to derive gates for. PR #30340 CI on 038f4be0 is unaffected by this round.", "dogfood_battery": "n/a — no `!` in the PR title and no (narrowing) arm", "mcp_calls": "0", "api_writes": "1 — this addendum comment on #30289 via post-stamped.mjs (relay, POST /repos/objectstack-ai/objectstack/issues/30289/comments). No push, no PR body update (nothing to describe yet).", "open_questions": [ { "question": "How should the v3 library land, given that the measured floor (~3330 changed lines) and the realistic size (~5000-6500) exceed the 3000-line human-merge threshold?", "options": [ "A. Split by landing order, every PR branched off main (not stacked). PR #30340 becomes `Part of #30289` and carries the five schema extensions + agentActions + their refusal tests + the 3 upgraded controls + deletion of the 6 retired cards (estimated ~2600-2900, re-measured before push; if it still crosses, the controls' upgrade moves out). Each mid-tier card then lands as its own PR (~500-900 lines each: card JSON + its own per-card truth-pin test file, so the six do not conflict), each `Blocked-by` #30340 and dispatchable in parallel once it merges; the last one closes #30289. Cost: 7 PRs and 7 reviews, but each is under the line and reviewable. Work: same total.", "B. One PR over 3000 via the AGENTS.md §7(c) path (an authorized APPROVED review, then the owning seat lands it). Cost: one ~5000-6500 line review for the maintainer-authorized approver, and nothing lands until all 6 mid-tier cards are done. Work: same total.", "C. Add a declarative seed generator to the card format (row templates x cycles over dates and owners), so 100+ rows cost a few lines and the whole round might fit one PR under 3000. Cost: a new format arm no ruling asked for, seeds no longer readable row by row, hand-verifying truths gets much harder (an oracle independent of the evaluator becomes impractical), and it expands the declared surface. Work: more than A." ], "recommendation": "A, on the four axes. Real business need: every piece serves a ruled card, and the schema arms are needed under every option, so landing them first unblocks parallel card authoring. Long-term soundness: small, reviewable, contract-first PRs; the per-card test file is the sustainable shape for a growing library. Hard-to-get-wrong for AI authors: the schema's strict refusals are in force before any mid-tier card is authored, so each card PR is judged by the hardened loader. No scope creep for the startup stage: A adds no format arm, while C adds a speculative one; B keeps one PR but piles everything onto one approver. The seat chooses; if A, the next patch round on this PR is the schema + controls slice and the six mid-tier cards are filed as sub-issues." } ], "security_findings": "none", "out_of_scope_findings": [] }objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Seat decision on the dev's fork (report comment 6658595859): split, option A
The dev stopped before pushing, as ordered: the v3 scope measured at a floor of ~3330 changed lines and a realistic ~5000–6500, over the 3000-line human-merge threshold. How work is sliced into PRs is task ordering, which the protocol does not escalate. Product and contract choices are not made here: option C would add a seed-generator format arm that no ruling asked for, and it is not taken.
Decision (A):
asOf+ relative windows, per-persona row scope, an unanswerable truth kind, declared write-operation outcomes);Fixes #30289stays: this card closes when the foundation lands.Blocked-by: #30340and dispatched in parallel once it merges: Scenario library: mid-tier card — 差旅报销 travel expense (8–10 问数 + 4–6 write operations, row-scoped personas) #30476 travel expense · Scenario library: mid-tier card — 订单到收款 order-to-cash (8–10 问数 + 4–6 write operations, row-scoped personas) #30477 order-to-cash · Scenario library: mid-tier card — 客服工单 service desk (8–10 问数 + 4–6 write operations, row-scoped personas) #30478 service desk · Scenario library: mid-tier card — 请假与假期余额 leave & balances (8–10 问数 + 4–6 write operations, row-scoped personas) #30479 leave & balances · Scenario library: mid-tier card — 采购申请到收货 requisition-to-receipt (8–10 问数 + 4–6 write operations, row-scoped personas) #30480 requisition-to-receipt · Scenario library: mid-tier card — 资产全生命周期 asset lifecycle (8–10 问数 + 4–6 write operations, row-scoped personas) #30481 asset lifecycle. Each lands its card plus its own truth-pin test file, under 3000 lines.objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Re-graded p2 → p1 on a maintainer ruling (Claude Code session 2026-11-05, recorded on objectstack-ai/cloud#3600), verbatim: 「都可以标记高优先级」, then on the seat's follow-up question the option 「p1 + 只交 6 张卡 (推荐)」. This card stays in the epic subtree (pipeline contract), so
pm:epicis unchanged. Note for triage: this is a direct maintainer re-grade. NORTH-STAR's "re-grade by the road table, never card by card" yields to the ruling by the skill's precedence (maintainer ruling > North Star).objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
os-dev-report
Addendum, patch round 1 (option A, foundation): done on head c1affd3. The PR body is not patched by the dev (os-dev.md); the replacement body for the seat to write is in the details block below.
{ "issue": 30289, "status": "done", "branch": "claude/issue-30289-scenario-card-schema", "pr": "https://github.com/objectstack-ai/objectstack/pull/30340", "session": "session_01Q4EjjPXwX1YfHyKJUh4agu", "premise_still_valid": true, "summary": "Patch round 1 (option A, foundation only) is delivered on head c1affd3f (merged with origin/main f506b79d as e5a49c33). It adds five schema extensions. (1) Relation paths on filter/groupBy, resolved through lookups declared in expect.lookups and the seed's lookup ids. (2) A card-level asOf plus a new in_window filter op (last_days N / calendar this|previous month|year), inclusive and deterministic. (3) A persona row scope ({object, field-or-path, value}); a question that lists personas also gets per-persona truths, returned as personaTruths, and the truth is `withheld` when the question reads an object or field that persona hides. (4) A truth kind `unanswerable` with a reason. (5) Every operation names actingAs and declares an outcome: applied post-state by id/where with values or absent, plus auditActor; refused with a code from the spec's closed ErrorCode catalogue; or pending_confirmation. The guardrail checks are held to those outcomes. Also: optional run.agentActions (>=3 actions, >=1 confirm-gated), and a counts band (>=5 questions and >=2 operations; below 8/4, countsReason is required). The expect shape is untouched. The 3 controls are upgraded: invoice_with_line_items 9 questions / 5 operations, project_with_tasks 8/5, crm_account_with_contacts 8/5. Each control has cross-object, per-persona and unanswerable questions (invoice also a this_month window), 5 operations (applied create, audited update, confirm-gated delete, refused VALIDATION_FAILED, refused PERMISSION_DENIED) and 3 agent actions. The 6 retired cards and their pins are git rm'd. The README documents the v3 library shape and the mid-tier floor, including the declarative floor. No mid-tier card is written. CI fix from the patch-round message: the refusal-code field is now `ErrorCode` from @objectstack/spec/api, read from the export and not restated, so the package depends on @objectstack/spec (workspace) with a vitest source alias. The controls expect VALIDATION_FAILED, the engine record validator's per-field code (packages/types validation-failure.ts, objectql VALIDATION_FAILED_CODE), instead of the generic VALIDATION_ERROR. CONFLICT noted, not resolved silently: the dispatch orders the dev to update the PR body via the relay, but .claude/agents/os-dev.md says the dev writes the body once at pr_create and never PATCHes it, with the seat writing later changes, and that file wins on conflict. So the body is not patched. The full replacement body is attached under the addendum comment's details block (four-backtick fence) for the seat to write. Size measured before every push: 2532 changed lines (+2529/-3, 20 files), under 3000.", "tests": "On head c1affd3f under os-verify-lock (VERDICT command-exit 0): pnpm --filter @objectstack/scenarios test -> Test Files 3 passed (3), Tests 59 passed (59). pnpm --filter @objectstack/spec build && pnpm --filter @objectstack/scenarios typecheck -> exit 0. library.test.ts: 25 truths + 10 persona truths, hand-worked. refusals.test.ts: 38 tabled refusal cases, each asserting exactly one refusal with file + exact path + message class, plus 2 controls and the id/basename and invalid-JSON cases. New-arm cases: relation paths 2 (run.questions[6].truth.groupBy, run.questions[5].truth.filter[0].field); windows 4 (run.questions[0].truth.filter[0].value x2, run.questions[0].truth.filter[0].field, asOf); persona scope/personas 2 (run.personas[0].scope[1].field, run.questions[4].personas[0]); unanswerable 1 (run.questions[7].truth.reason); outcomes 5 (run.operations[0].outcome, run.operations[0].actingAs, run.operations[3].outcome.code, run.operations[0].outcome.state[0].id, run.operations[1].outcome.state[0].id); guardrail-outcome 2 (run.guardrails[2].operation, run.guardrails[3].operation); agentActions 2 (run.agentActions, run.agentActions[0].object); countsReason 1 (run.countsReason). Ablations, one per new arm, via scripts/ablation-replace.mjs (tests import ../src; spec aliased to source, so no dist leg), each restored byte-identical (git hash-object == HEAD blob), git diff HEAD empty after: relation_paths | exit 1 | landed True | blob 73dcb24d->3fb8af58 | 32 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: ['loads every card with no refusal', 'holds exactly the three control cards (the mid-tier cards land on their own)'] ;; time_windows | exit 1 | landed True | blob d562f0d3->9254d982 | 2 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: [\"invoice_with_line_items: every question's computed truth is the hand-worked one\", 'computes windows against asOf, inclusive, across month and year edges'] ;; persona_row_scope | exit 1 | landed True | blob 73dcb24d->c801f0c3 | 3 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: [\"invoice_with_line_items: every question's computed truth is the hand-worked one\", \"project_with_tasks: every question's computed truth is the hand-worked one\"] ;; persona_withheld | exit 1 | landed True | blob 73dcb24d->39a838a3 | 3 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: [\"invoice_with_line_items: every question's computed truth is the hand-worked one\", \"project_with_tasks: every question's computed truth is the hand-worked one\"] ;; unanswerable | exit 1 | landed True | blob d562f0d3->f9a664ff | 4 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: [\"invoice_with_line_items: every question's computed truth is the hand-worked one\", \"project_with_tasks: every question's computed truth is the hand-worked one\"] ;; outcome_guardrail_match | exit 1 | landed True | blob 73dcb24d->d93bd217 | 1 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: ['refuses a confirmation check on an operation that does not wait'] ;; outcome_post_state_id | exit 1 | landed True | blob 73dcb24d->f1d67cba | 1 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: ['refuses a post-state naming a created row by an id the seed lacks'] ;; agent_actions_confirm | exit 1 | landed True | blob 73dcb24d->58d94865 | 1 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: ['refuses agent actions with none confirm-gated'] ;; counts_reason | exit 1 | landed True | blob 73dcb24d->313cb397 | 1 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: ['refuses fewer than 8 questions without a stated reason'] ;; refusal_code_catalogue | exit 1 | landed True | blob 47f63c7d->163daac4 | 1 ⎯⎯⎯⎯⎯⎯⎯ | restored True | red: ['refuses a refusal code outside the platform error-code catalogue'] ;; git diff HEAD empty: True", "gates": "dispatch-gates --commands on c1affd3f (20 paths; changed lines 2532 vs threshold 3000: under), each run with the exit recorded before any pipe; `--ran` -> '92 derived famil(ies) accounted for — 88 run, 4 NOT-MEASURED (3 DERIVED from a recorded exit 3, 1 claimed)', exit 0. All 88 run exit 0, including check:pm-dispatch-gates (rerun without the wrapper timeout, exit 0), check:cross-package-test-inputs, check:ci-filter-parity, check:test-source-alias and check:undeclared-dep-imports. check:error-code-casing, the CI red on c3a257d5: reproduced locally exit 1 ('packages/qa/scenarios/test/refusals.test.ts:83 validation_error (assignment)'); exit 0 on c1affd3f ('no unlisted lowercase error codes in 11258 scanned file(s)'); no EXEMPT_FILES entry. NOT MEASURED: exit 3 PREREQUISITE NOT MET (no dist for those packages) check:dual-build-cjs-loads, check:lean-entry-closure, @objectstack/lint check:doc-formula-expressions; exit 124 under the 420 s wrapper check:type-check-debt (CI-owned). CI on c1affd3f at report time: 58 check-runs success, the rest skipped, none red or pending.", "dogfood_battery": "n/a — no `!` in the PR title and no (narrowing) arm", "mcp_calls": "0", "api_writes": "1 — this addendum comment on #30289 via post-stamped.mjs (relay, POST /repos/objectstack-ai/objectstack/issues/30289/comments). The PR body was NOT patched (os-dev.md rule: the dev writes the body once; the seat writes later changes). The replacement body is attached in this comment.", "open_questions": [], "security_findings": "none", "out_of_scope_findings": [ "carrier: mid-tier sub-issues #30476-#30481 · noted, not filed: the existing rule that every persona hides at least one object or field means a card has no all-seeing admin persona, and operations act as scoped personas. If a mid-tier card needs an admin role, relaxing that rule is a card-author decision for those sub-issues; it is not made here.", "carrier: none · noted, not filed: write_validation outcomes assume the runtime answers VALIDATION_FAILED (record validator) for a missing required field or a malformed email. This is grounded in the engine's ValidationError code, not in a booted run. The local runner card measures it against a live app." ] }Replacement PR body for #30340 (seat to write)
objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Note for the mid-tier card authors (#30476–#30481, now in the
domain:devxlane), from the PR #30340 review. The schema keeps the rule that every persona hides at least one object or field, so a card has no all-seeing admin persona and operations act as scoped personas. If a mid-tier card needs an all-seeing role (e.g. finance over all invoices), relaxing that rule is a schema change: raise it on that card's PR as a stated deviation with its reason, not a silent edit. Also:write_validationoutcomes assume the record validator answersVALIDATION_FAILED; the runner #30292 measures that live.objectstack-fleet commentedon Nov 5, 2026
epic:#3600#1 · hotlong · session_01Q4EjjPXwX1YfHyKJUh4agu
Landed: PR #30340 merged as
58085ffd, closing this card (Fixes #30289).pm:dispatchedremoved with the close. Unblocked by this landing: the six mid-tier cards #30476–#30481 (released into the devx lane'spm:queuein this round) and T2b #30291 (both itsBlocked-by:now closed; dispatched by the epic PM).