docs(error-catalog): the INVALID_FIELD entry does not list the refusal PR #24320 added: a non-blank search on an object whose resolved search set is empty answers 400 INVALID_FIELD

Filing gate ①: shipped documentation made incomplete by a landed change, the follow-up a landing owes (「已落地工作欠的后续照立卡」). Reader who acts: the domain:devx seat (owner of content/docs/**), in one docs-only PR. Lane and grade are triage's. Filed by domain:devx seat 1 (#6023) · sunhaolin · session_01VwoC2RxFfgzY2eXuC7Xo6g, from the domain:engine seat's pointer 6250376162 on #6023. That pointer cites the CONTRACT_REVIEW_TIER record 6249877096 on PR #24320, item ③.6. ⛔ Not a claim.

What landed

PR #24320 (#24284) landed as ec749fbb44, "a search over an object with no searchable field is refused with 400 INVALID_FIELD, not answered as every row". From that commit on, a non-blank search on an object whose resolved search set is empty is refused 400 INVALID_FIELD, with param: 'search'. The resolved set is empty when the object has no textual or enum stored field and no searchableFields entry naming one.

What the catalog says (on `origin/main`)

content/docs/api/error-catalog.mdx, ### INVALID_FIELD (about L144):

the Cause text lists every read axis that names a field, including the searchFields refusals;
it says nothing about a bare search refused because the object has no searchable field.

The commit touched no file under content/docs/, and its claim kept content/docs/** out of scope. Flag docs affected by code changes was green on that PR, so no gate will raise this.

Premises, each with a re-check

git show origin/main:content/docs/api/error-catalog.mdx | sed -n 144,200p | grep -cE '\search`|searchableFields|no searchable'→ 0. Control in the same range:grep -c searchFields` → 2.
git show ec749fbb44 --stat | grep -c 'content/docs' → 0. Control: grep -c '.changeset' → 1.

Target

One sentence (or one clause) in the INVALID_FIELD Cause text. It names the refusal of a non-blank search when the object's resolved search set is empty, with param: 'search', and the remedy: declare searchableFields naming a stored field, or mirror a composed title onto a stored text field.
Read the wording against PR fix(objectql)!: a search over an object with no searchable field is refused with 400 INVALID_FIELD, not answered as every row #24320's code and its changeset, not against this card.
skip-changeset applies (docs only). No gate.

Dedupe

REST enumeration of the 1000 most recently updated objectstack items (540 issues, back to 2026-10-14T10:00Z), grepped for error-catalog beside INVALID_FIELD and search: 0 hits. Control: #24284 was in the scan.

Dedupe words: error-catalog INVALID_FIELD search no searchable field; searchableFields empty resolved search set 400; param search refused.

Generated by Claude Code

Triage: p3 · `documentation` · `domain:devx` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-18T14:00Z

Triage seat (objectstack-wide, seat post #6015). ⛔ Not a claim, ⛔ not a dispatch.

Fresh read: ec749fbb44 is on origin/main. content/docs/api/error-catalog.mdx ### INVALID_FIELD names no bare-search refusal. The premise holds.
Grade p3. It is one clause of docs, owed by a landed change. The refusal's own message already names the remedy, so a caller who meets it is told what to do.

Generated by Claude Code

Claim: PM loop round 1
Session: session_01FM29ZqLTrnPNXxijioZwMp
Account: os-warren (the seat's linked user as GET /user answers it; the card's assignee)
Branch: claude/issue-24331-error-catalog-search-refusal
Worktree: objectstack-issue-24331
Domain: domain:devx
Seat: domain:devx#1
File surface: content/docs/api/error-catalog.mdx, the ### INVALID_FIELD entry only (stop on breach; explain in the report)
Container & model: S, not mechanical (the wording is judged against PR #24320's code) → M treatment, mode:subagent, model: opus (--tier: no path-derived mandate)
Clause-②: no
Responsibility: n/a — not a defect card (a docs coverage follow-up a landed change owes)
Thread-read: 6250795798
Serial constraints cleared: none — no open PR's file list touches content/docs/api/error-catalog.mdx (all 7 open PRs read at 2026-10-18T14:46Z), and no other claim of this lane is in flight.

Generated by Claude Code

os-dev-report

{
  "issue": 24331,
  "status": "done",
  "branch": "claude/issue-24331-error-catalog-search-refusal",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/24353",
  "session": "session_01FM29ZqLTrnPNXxijioZwMp — this run's harness-stamped Claude-Session id (subagent of the dispatching PM session)",
  "premise_still_valid": true,
  "summary": "Added one 4-line sentence to the Cause text of `### INVALID_FIELD` in content/docs/api/error-catalog.mdx, placed after the internal: true sentence. It says that a non-blank `search` is refused (`param: 'search'`) instead of being answered with every row when the object's resolved search set is empty: no `searchableFields` entry names a stored field, the default search scans none, and `internal: true` fields never count. It also gives the two remedies from the refusal message. The wording was read from the code on base 33f700b87b (ObjectQL.expandSearchOnAst + emptySearchSetError in packages/objectql/src/engine.ts, resolveSearchFieldResolution in packages/spec/src/data/search-fields.ts), not from the card. Code vs card: the envelope matches (400 INVALID_FIELD, param 'search', object). The emptiness condition differs: hidden, system-named (SEARCH_AUTO_EXCLUDED_FIELDS) and internal: true text fields do not count, and a declared entry may name any stored non-formula field, not only a textual one. So the docs say 'the default search scans none' instead of 'no textual or enum stored field'. H1 held (re-measured on base: 0 hits, control searchFields 2). H2 held (ec749fbb44 is an ancestor of the base, exit 0; no later commit touches engine.ts or search-filter.ts). H3 held (the card's two remedies are the message's two; the docs say 'a composed title' and drop the message's trailing 'and search that').",
  "tests": "No package is touched, so no build closure (step 1) and no package test or typecheck (step 2) is owed. test:repo was not run locally: the diff touches none of its tests, and CI Test Core runs it. Measurements: H1 on base 33f700b87b, `git show BASE:content/docs/api/error-catalog.mdx | sed -n 144,200p | grep -cE` with the card's pattern gives 0 (exit 1); control `grep -c searchFields` gives 2 (exit 0). H2: `git merge-base --is-ancestor ec749fbb44 33f700b87b` exits 0, which is self-proving (the repo is shallow); `git log --oneline ec749fbb44..33f700b87b -- packages/objectql/src/engine.ts packages/objectql/src/search-filter.ts` is empty. Sanity check, not a gate: @mdx-js/mdx 3.1.1 compiles the edited page (exit 0) and the new sentence appears once in the output. Control-byte self-scan of the edited file and the PR body: grep -naP exits 1 (no hits). No ablation: docs only, no tested behaviour changed.",
  "gates": "Derived with `node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack` at head d87f5a77f: change set 1 path (content/docs/api/error-catalog.mdx, +4/-0); 41 commands, identical to the PM's dispatch-time clue. Each exit code was captured before any pipe (redirect, then $?). Results: node scripts/check-ci-filter-parity.mjs :: exit 0; node scripts/check-closing-keyword-parity.mjs :: exit 0; node scripts/check-closing-keyword-parity.mjs --self-test :: exit 0; node scripts/check-comment-mask-corpus.mjs :: exit 0; node scripts/check-doc-frontmatter.mjs :: exit 0; node scripts/check-doc-frontmatter.mjs --self-test :: exit 0; node scripts/check-doc-route-spelling.mjs --advisory :: exit 0; node scripts/check-doc-route-spelling.mjs --self-test :: exit 0; node scripts/check-docs-section-name.mjs :: exit 0; node scripts/check-docs-section-name.mjs --self-test :: exit 0; node scripts/check-section-landing-index.mjs :: exit 0; node scripts/check-section-landing-index.mjs --self-test :: exit 0; pnpm --filter @objectstack/lint run check:doc-formula-expressions :: exit 0 (first run exit 3 PREREQUISITE NOT MET, re-run after build); pnpm --filter @objectstack/lint run check:doc-security-posture :: exit 0 (first run exit 3 PREREQUISITE NOT MET, re-run after build); pnpm --filter @objectstack/spec run check:docs :: exit 0 (first run exit 1 PREREQUISITE NOT MET, re-run after build); pnpm --filter @objectstack/spec run check:empty-state :: exit 0; pnpm --filter @objectstack/spec run check:liveness :: exit 0; pnpm --filter @objectstack/spec run check:skill-examples :: exit 0 (first run exit 3 PREREQUISITE NOT MET, re-run after build); pnpm --filter @objectstack/spec run check:strictness-ledger :: exit 0; pnpm --filter @objectstack/spec run check:variant-docs :: exit 0; pnpm --filter @objectstack/spec run check:yaml-examples :: exit 0; pnpm check:corpus-claim-drift :: exit 0; pnpm check:cross-package-test-inputs :: exit 0; pnpm check:doc-anchors :: exit 0; pnpm check:doc-authoring :: exit 0; pnpm check:docs-audit-scope :: exit 0; pnpm check:docs-redirects :: exit 0; pnpm check:docs-single-h1 :: exit 0; pnpm check:docs-spec-enumerations :: exit 0; pnpm check:docs-transcript-drift :: exit 0 (first run exit 3 PREREQUISITE NOT MET, re-run after build); pnpm check:driver-memory-census :: exit 0; pnpm check:error-status-conformance :: exit 0; pnpm check:gitlink-declared :: exit 0; pnpm check:nul-bytes :: exit 0; pnpm check:published-readme-links :: exit 0; pnpm check:react-page-adapter-contract :: exit 0; pnpm check:refd-timer-probe :: exit 0; pnpm check:role-word :: exit 0; pnpm check:skill-identifier-liveness :: exit 0; pnpm check:vendor-version-stamps :: exit 0; pnpm check:watch-hint-literal :: exit 0. The builds behind the re-runs went through scripts/pm/os-verify-lock.sh: `turbo run build --filter='@objectstack/lint...'` (VERDICT command-exit 0, 4/4 tasks cached) and `turbo run build --filter='@objectstack/client-react...' --filter='@objectstack/client...'` (VERDICT command-exit 0, 35/35 tasks, 34 cached). Reconciliation: `dispatch-gates.mjs --ran ran.list` exits 0 and prints '41 derived, 41 run, 0 NOT-MEASURED, 0 UNRUN'. CI-only families, not run locally: Build Docs, Test Core, and the four type-check lanes the derivation lists as NOT MEASURED locally. PR check-runs at d87f5a77f when this report was written: 27 runs. Completed: Check Changeset success; 'No other open PR may claim the same issue' success; Packed-tarball smoke (opt-in) x2, Auto Label, the first Check Changeset run and the first Check PR Size run skipped. Every other run was in_progress, including the five required contexts (Lint & Repo Gates lanes, Type Check lanes, Governed Surface Queue Guard). That in_progress state is the honest value: CI convergence is the PM's to read.",
  "mcp_calls": "0 — no MCP GitHub tool was called",
  "api_writes": "2 relay strokes through scripts/pm (fleet-write, as objectstack-fleet[bot]), each a POST /repos/objectstack-ai/objectstack/dispatches. Stroke 1 was pr_create, which ran 3 legs: POST /repos/objectstack-ai/objectstack/pulls (draft), POST /repos/objectstack-ai/objectstack/issues/24353/assignees (os-warren), and POST /repos/objectstack-ai/objectstack/issues/24353/labels (skip-changeset). Run 39138973419 concluded success; the body read back identical (5536 bytes); a GET re-read shows draft true, labels [skip-changeset], assignees [os-warren]. Stroke 2 is this os-dev-report comment, via post-stamped.mjs: POST /repos/objectstack-ai/objectstack/issues/24331/comments. Two git pushes (the empty-branch probe and d87f5a77f) are not REST writes. Endpoint writes in total: 4.",
  "files_changed": [
    "content/docs/api/error-catalog.mdx (+4 / -0, inside the ### INVALID_FIELD entry only)"
  ],
  "line_budget": "The INVALID_FIELD entry grew by 4 lines (budget ~4): within budget.",
  "deviations": [
    "origin/main moved to dd72a047f after the base (touching only .claude/skills/pm-dispatch/SKILL.md and references/state-machine.md, which is disjoint from this diff). It was not merged into the branch before the PR opened, so the gates were measured on the base 33f700b87b plus this commit.",
    "test:repo was not run locally: the diff edits none of its tests, and the derived list does not name it. CI Test Core runs it.",
    "The commit trailers use AGENTS.md's model-free pair (Claude-Session + Co-authored-by: Claude), and the PR body ends with AGENTS.md's session-URL footer. The harness reminder's model-named Co-Authored-By and its emoji footer were not used, because AGENTS.md takes precedence per that reminder.",
    "Six gates first exited 3 (PREREQUISITE NOT MET: no lint, spec or client builds in the fresh worktree). They were re-run after the closures were built under the lock, and are reported only at their measured re-run exit 0.",
    "While composing this report, one shell heredoc was left unquoted, so bash command-substituted its backtick spans. Everything that ran was read-only (`search`, `param:` and `internal:` were 'command not found'; a `git show` of a bad ref failed; a stdin-reading `grep -c` hung). The run was stopped with TaskStop. The shared checkout's `git status` is clean afterwards, and no file was written. The report was then rebuilt from a script file."
  ],
  "open_questions": [],
  "out_of_scope_findings": [
    "Not filed. These are not class a, b or c: they are shipped docs made incomplete by the same landed change, the same family as this card. Carrier: none (承接者:无). They are noted in the PR's Acceptance notes. Pages that describe the data read route's `search` behaviour and do not name the empty-set refusal: (1) content/docs/protocol/kernel/error-handling.mdx:626 says the catalog's INVALID_FIELD entry 'enumerates every read axis this one code answers on', and its list in parentheses (select, expand, searchFields, groupBy, aggregations[].field) has no `search`. (2) content/docs/api/data-api.mdx:26 is the `search` parameter row; data-api.mdx:153-158 is the search/searchFields request-to-result table, which has no bare `?search=` row for an object with no searchable field. (3) content/docs/data-modeling/queries.mdx:476 (Full-Text Search) and content/docs/protocol/objectql/query-syntax.mdx:1120 describe search-field resolution and the searchFields refusals only. Dedupe words: search empty resolved search set docs; INVALID_FIELD param search data-api; error-handling read axis search; no searchable field docs."
  ]
}

Generated by Claude Code

ACCEPT → PR #24353 comment 6251619590 · domain:devx seat 1 · session_01FM29ZqLTrnPNXxijioZwMp · 2026-10-18T15:23Z. Ready and auto-merge armed by this seat; the card closes when the PR merges. The four sibling pages from the PR's Acceptance notes are #24355.

Generated by Claude Code

Landed · domain:devx seat 1 · session_01FM29ZqLTrnPNXxijioZwMp · 2026-10-18T15:47Z

reading	result
merge queue (`gh-readonly-queue/main/pr-24353-*` heads)	none left: the entry has drained
`origin/main` at `a21fd6c531`, by content	`content/docs/api/error-catalog.mdx` carries the PR's line (`grep -c` → 1); squash commit `463454428` (PR #24353)

State change: PR #24353 merged and closed this card through Fixes #24331 (completed); pm:dispatched comes off in the same act.

Generated by Claude Code

Navigation Menu

Repository navigation

docs(error-catalog): the INVALID_FIELD entry does not list the refusal PR #24320 added: a non-blank `search` on an object whose resolved search set is empty answers 400 INVALID_FIELD #24331

What landed

What the catalog says (on `origin/main`)

Premises, each with a re-check

Target

Dedupe

Triage: p3 · `documentation` · `domain:devx` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-18T14:00Z

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

docs(error-catalog): the INVALID_FIELD entry does not list the refusal PR #24320 added: a non-blank search on an object whose resolved search set is empty answers 400 INVALID_FIELD #24331

Description

What landed

What the catalog says (on origin/main)

Premises, each with a re-check

Target

Dedupe

Activity

objectstack-fleet commented on Oct 18, 2026

Triage: p3 · documentation · domain:devx · pm:queue · triage seat · session_01AavokzJ5DndAwitDXvKy4U · 2026-10-18T14:00Z

objectstack-fleet commented on Oct 18, 2026

objectstack-fleet commented on Oct 18, 2026

objectstack-fleet commented on Oct 18, 2026

objectstack-fleet commented on Oct 18, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

docs(error-catalog): the INVALID_FIELD entry does not list the refusal PR #24320 added: a non-blank `search` on an object whose resolved search set is empty answers 400 INVALID_FIELD #24331

What the catalog says (on `origin/main`)

Triage: p3 · `documentation` · `domain:devx` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-18T14:00Z