Lead and case detail pages: show the sections a user is expected to fill even while empty (hideEmpty: false), the way #1211 did for the opportunity page #2003

Activity

1. objectstack-fleet

   added

   enhancementNew feature or requestNew feature or request

   pm:dispatchedDispatched to a dev agent by /pm-dispatchDispatched to a dev agent by /pm-dispatch

   priority:p3

   on Oct 3, 2026
2. objectstack-fleet

   assigned

   hotlong

   on Oct 3, 2026
3. objectstack-fleet commented on Oct 3, 2026

   

   objectstack-fleetbot

   on Oct 3, 2026

   ContributorAuthor

   More actions

   Claim: PM loop round R73
   Session: session_01ER8ntXZhYebyQ66aXWdjfT
   Account: hotlong (the seat's linked user as GET /user answers it; always the card's assignee)
   Branch: claude/issue-2003-lead-case-empty-groups
   Worktree: hotcrm-issue-2003
   Domain: repo:hotcrm (single-lane repo, no domain:* taxonomy)
   Seat: repo:hotcrm#1
   File surface: src/sales/pages/lead_detail.page.ts, src/service/pages/case_detail.page.ts (the hideEmpty key on their { group } sections only), pins under test/ that read them, one .changeset/2003-*.md (stop on breach; explain in the report)
   Container & model: S, mode:subagent, model: default (dispatch-gates --tier --repo objectstack-ai/hotcrm answers "no path-derived mandate … the tier stays the PM's per-card judgment call")
   Clause-②: no. This is app metadata; it touches no published schema and no accept set.
   Thread-read: none
   Serial constraints cleared: three cards are dispatched together under the maintainer's 「决裁的三张同意」: #1992 (src/service/{objects/case.*,reports,datasets,dashboards,views,data}), #2003 (src/sales/pages/lead_detail.page.ts, src/service/pages/case_detail.page.ts) and #2004 (src/sales/objects/opportunity.hook.ts, the 立项 gate test, the qualification docs). Their files are disjoint. #1992 shrinks the sla group that #2003 decides hideEmpty for; each dev is told to measure on its own tree and name that coupling. No open PR exists.

   ---

   Generated by Claude Code
4. objectstack-fleet commented on Oct 3, 2026

   

   objectstack-fleetbot

   on Oct 3, 2026

   ContributorAuthor

   More actions

   os-dev-report

   {
    "issue": 2003,
    "status": "done",
    "branch": "claude/issue-2003-lead-case-empty-groups",
    "pr": "https://github.com/objectstack-ai/hotcrm/pull/2005",
    "session": "session_01ER8ntXZhYebyQ66aXWdjfT — subagent, so this is the parent seat's id",
    "premise_still_valid": true,
    "summary": "The premise held: on fresh records under 17.6.0, a lead lost 5 of its 10 group sections (contact_info, assignment, address, duplicates, additional) and a case lost 3 of 6 (origin, sla, resolution). PR #2005 (draft, assigned to hotlong) adds `hideEmpty: false` to four `{ group }` sections: lead contact_info, address and additional, and case origin. It also adds a patch changeset. No fieldGroups change. Chromium before/after on fresh records: the lead goes from 5 sections to 8, the case from 3 to 4, and each page shows exactly the chosen sections. Each group was decided from its members, and two decisions differ from the card's initial reading. (1) Lead `additional` is ON: every member (annual_revenue, number_of_employees, description, notes) is typed by a rep, none is readonly or written by any hook or flow, and the default create form asks for all four; it is the lead's equivalent of the opportunity `notes` group from #1211, and the revert is one key. (2) Case `sla` is OFF. I decided it on the six members other than created_date, as the coordinator directed (#1992 is blocked upstream on objectstack#21646). priority, sla_due_date and is_sla_violated are in the strip. The other three are system stamps (closed_date and resolution_time_hours from case.hook at close, first_response_date from event.hook.ts). The decision holds whether or not created_date is retired. Lead `assignment` stays off because the key is inert: owner_id is in the strip, the derived list is empty, and the probe confirms nothing renders even with the key. Lead `duplicates` (written by the lead_duplicate_check hook) and case `resolution` (collected by close_case) stay hidden.",
    "tests": "pnpm verify: `OS_VERIFY_LOCK_SLOT=hotcrm-issue-2003 bash /home/user/objectstack/scripts/pm/os-verify-lock.sh -c 'NODE_OPTIONS=--max-old-space-size=4096 pnpm verify'` at HEAD 6e68fa1e (git rev-parse --short HEAD read before and after the same run) gave `os-verify-lock: VERDICT command-exit 0 · held the lock 177s`. Chain: `✓ Validation passed` · tsc --noEmit clean · lint `1 warning(s), 18 suggestion(s)` · `✓ i18n lint gate: 0 i18n/missing-* issues` · `✓ source hygiene clean` · `✓ source token ratchet clean` · `✓ Build complete` (9 author-time warnings) · `Test Files 177 passed (177)`, `Tests 3794 passed | 1 skipped (3795)`. Lint, BASE vs HEAD pages: identical output, 1 warning and 18 suggestions both times. The pages were swapped to BASE with a trap restore; after restore `git diff HEAD` was empty and both blob hashes equalled the HEAD blobs (e45cd08b95ce, 8d73fa80ea01). Author-time warning list identical, 9 to 9. Token ratchet (node scripts/check-source-token-ratchet.mjs) before and after, src/sales rows verbatim: before `interaction layer 25 3,373 111,870 27,968 31,000` and `authored total 117 10,417 399,717 99,929 107,000`; after `interaction layer 25 3,373 111,924 27,981 31,000` and `authored total 117 10,417 399,771 99,943 107,000`. Business semantics unchanged at 56,418. src/service interaction layer went from 5,592 to 5,597. No ceiling changed. Legality: RecordDetailsProps (=== ComponentPropsMap['record:details']).safeParse succeeds on both pages and keeps hideEmpty false on contact_info, address, additional and origin. Three controls per page are refused: hideEmpty 'no' (invalid_type), hideEmptyX (unrecognized_keys), collapsible beside group (custom). PageSchema.safeParse accepts all three controls, so it is not the proof. Browser (Chromium /opt/pw-browsers/chromium, port 4821, a fresh OS_HOME per boot, served record:details sections EQUAL to dist/objectstack.json before every probe): before 94668373: lead 5 cards, case 3 cards. After 1ebe784c: lead 8 cards (+ Contact Information with 2 empty rows, + Address collapsed with 1, + Additional Info collapsed with 4); case 4 cards (+ Origin & Routing with 1 empty row). Probe-only artifact with the key on all four groups left off: lead 9 cards (assignment still absent; duplicates draws 4 empty rows); case 6 cards (sla draws 4 empty rows, resolution 2). The first probe boot was VOIDED: `--artifact` run from the worktree cwd served the worktree's dist instead (served != probe on both pages, the objectstack#21501 shape). The rerun from a cwd with no config gave EQUAL. NOT MEASURED: a seeded record with values in the browser, reason: covered by reading the console bundle (31971ff1), where hideEmpty is read only in the all-empty branch, and by #1211's measurement on the same bundle. Control-byte self-scan of the three changed files and the PR body: 0 hits. No test added; no test reads hideEmpty. Files changed: src/sales/pages/lead_detail.page.ts, src/service/pages/case_detail.page.ts, .changeset/2003-lead-case-empty-groups.md. Deviations: (a) the two departures from the initial reading, above; (b) commit trailers are the model-free pair (Co-authored-by Claude, Claude-Session), per the dev contract, rather than the harness reminder's model-named line. Cleanup: every server I started was stopped by its own process group (1815, 6138, 7240, 7869); the worktree was removed without --force.",
    "mcp_calls": "0",
    "api_writes": "3 relay writes as objectstack-fleet[bot], each one POST /repos/objectstack-ai/objectstack/dispatches executed by fleet-write: (1) pr_create, POST /repos/objectstack-ai/hotcrm/pulls (draft forced), giving #2005; read-back byte-identical, 10979 bytes. (2) label-write assign, POST /repos/objectstack-ai/hotcrm/issues/2005/assignees with hotlong; read-back MATCHES. (3) this os-dev-report comment, POST /repos/objectstack-ai/hotcrm/issues/2003/comments. Plus git push to the branch (not REST). The PR's `metadata` label was not written by me (labeler).",
    "open_questions": [],
    "out_of_scope_findings": [
     "carrier: none (承接者:无) · noted in PR Acceptance notes, not filed · the token ratchet prints a shrink-only hint to re-anchor the src/sales interaction-layer ceiling to ~30,000 and the src/service one to ~6,000; that is outside this card's file surface and not a defect · dedupe words: token ratchet re-anchor ceiling headroom",
     "carrier: objectstack-ai/objectstack#21501 (existing card) · noted, not filed · reproduced it once more: `objectstack dev --artifact X` run from a cwd that holds objectstack.config.ts serves that cwd's dist/objectstack.json, not X; the served-metadata check caught it · dedupe words: artifact flag overridden cwd config served stale"
    ]
   }

   ---

   Generated by Claude Code
5. objectstack-fleet commented on Oct 3, 2026

   

   objectstack-fleetbot

   on Oct 3, 2026

   ContributorAuthor

   More actions

   ACCEPT: PR #2005 at 6e68fa1e. Lead and case detail pages show the sections a user is expected to fill. repo:hotcrm seat, session_01ER8ntXZhYebyQ66aXWdjfT, 2026-10-03T23:38Z. Report: the dev's os-dev-report 5974648404.

   Probes taken by the seat itself:

   • Ancestry and checks: main 94668373 is an ancestor of 6e68fa1e. CI 9/9 success. Governed check: 0 of 3 paths, NOT governed; 56 lines.
   • Read line by line: exactly four hideEmpty: false keys:
     • lead: contact_info, address, additional;
     • case: origin.
       Each has its reason beside it, plus a patch changeset. No fieldGroups change.

   Per-group decisions, accepted on their stated grounds:

   • Lead additional is ON (a departure from the initial reading). Every member, annual_revenue, number_of_employees, description and notes, is typed by a rep, none is written by a hook, and the create form asks for all four. It is the lead's counterpart of the opportunity's notes from The opportunity's 详情 tab renders two fields out of the fourteen it authors — highlight fields are dropped and the sections left empty disappear #1211.
   • Case sla is OFF (a departure from the initial reading). Decided on its six members other than created_date, as the seat directed while crm_case.created_date has no writer — a user-created case stores null and drops out of the "Cases Opened by Priority × Day" report and the service dashboard's date range (#575 B2's shape, on the case object) #1992 waits on A seed row's explicit created_at is overwritten with the boot instant on INSERT (seed context sets no preserveAudit), yet written on the upsert UPDATE of a later boot — seeds cannot backdate creation time consistently objectstack#21646:
     • priority, sla_due_date and is_sla_violated sit in the strip;
     • closed_date, resolution_time_hours and first_response_date are system stamps.
       The decision holds whether or not created_date retires.
   • Lead assignment is not keyed: it is inert. owner_id is in the strip, and the probe shows nothing renders even with the key.
   • Lead duplicates (hook-written) and case resolution (collected at close) stay hidden.

   The dev's measurements, accepted:

   • Chromium, fresh 17.6.0 boots (served sections checked equal to the artifact; one boot was voided under objectstack dev -a PATH prints Artifact: PATH but serves dist/objectstack.json, and objectstack start --artifact PATH run beside an objectstack.config.ts serves the config — the explicit artifact flag loses to the cwd objectstack#21501 and rerun): the lead goes from 5 sections to 8, the case from 3 to 4, exactly the chosen ones.
   • Legality: RecordDetailsProps.safeParse accepts both pages and refuses three controls per page.
   • Gates: pnpm verify VERDICT command-exit 0 (177 files). Tokens: src/sales interaction +13, no ceiling moved.

   Landing:

   • Authority: the maintainer's 「决裁的三张同意」 (option A for this item), and 「授权你执行pr合并」.
   • Path: ready → auto-merge → queue.

   ---

   Generated by Claude Code
6. objectstack-fleet

   closed this as completedin #2005on Oct 3, 2026
7. objectstack-fleet

   removed

   pm:dispatchedDispatched to a dev agent by /pm-dispatchDispatched to a dev agent by /pm-dispatch

   on Oct 3, 2026
8. objectstack-fleet

   unassigned

   hotlong

   on Oct 3, 2026