[finding] content/docs/protocol/kernel/error-handling.mdx documents DUPLICATE_RECORD as a 409 HTTP response code, while every route reports the conflict as UNIQUE_VIOLATION

Observation recorded while implementing #14723 (maintainer ruling A: a unique-constraint refusal has ONE wire spelling on every route, UNIQUE_VIOLATION; the engine's thrown DuplicateRecordError.code stays DUPLICATE_RECORD in-process). Filed unassigned; not addressed by the #14723 PR, whose surface is code, pins and changesets.

What the page says

content/docs/protocol/kernel/error-handling.mdx, section ## Error Codes / ### Resource Errors, at line 349 on origin/main 2ed6be649:

#### DUPLICATE_RECORD — "HTTP Status: 409", "Meaning: Resource with unique constraint already exists", with an example HTTP body whose error.code is "DUPLICATE_RECORD" (and a details object carrying resource / constraint / field / value).
Line 950, in the rate-limiting section: // Response: "DUPLICATE_RECORD" or "VALIDATION_ERROR" for a registration attempt.

What the platform does

content/docs/protocol/kernel/http-protocol.mdx lines 759 and 775 document the 409 constraint-violation body as "code": "UNIQUE_VIOLATION" — the spelling the ruling on rest/metadata-protocol: one unique conflict, two wire spellings on ONE route — a batch/import ROW reports DUPLICATE_RECORD while the whole-request failure on the same route reports UNIQUE_VIOLATION #14723 names as the published one.
Every REST route answers a unique-constraint refusal as UNIQUE_VIOLATION: the single-record door (唯一约束冲突没有单一判别谓词:仓内四套各自为政的方言词表,REST 的 409 映射漏掉 MySQL(Duplicate entry 落成 500 INTERNAL_ERROR) #6250 / REST: the UNIQUE_VIOLATION 409 message is hard-coded English and carries no field — and it is less informative than the bulk path's own message for the same constraint #7821 / mapDataError has no DUPLICATE_RECORD arm: the engine's insert conflict envelope reaches the wire through the generic passthrough, dropping the field key and the user-facing conflict sentence #14389), the whole-request bulk / import / metadata / UI doors (Routes exiting through handleRouteError answer the engine's DUPLICATE_RECORD envelope from resolveErrorResponse's .status passthrough — field and object dropped, mapDataError's structured 409 arms never consulted #14541), and after rest/metadata-protocol: one unique conflict, two wire spellings on ONE route — a batch/import ROW reports DUPLICATE_RECORD while the whole-request failure on the same route reports UNIQUE_VIOLATION #14723 the per-row reports of POST /api/v1/data/:object/batch and the import runner.
DUPLICATE_RECORD is the engine's in-process identity (a hook or a flow node catching DuplicateRecordError sees it — content/docs/automation/flows.mdx line 1124 documents that correctly), never an HTTP body's code.

So the error-handling page describes a wire body no route emits, and the example's details.value echoes the offending value, which the platform sentence deliberately withholds (the driver's error stays on cause).

Not measured here

Whether the { "success": false, "error": { ... } } envelope shape on that page matches today's ADR-0112 body at all — the whole section may predate the current envelope; this card records only the code spelling.
Whether content/docs/api/error-catalog.mdx line 355 (### DUPLICATE_RECORD, "A record with the same unique key already exists") needs a note distinguishing the in-process code from the wire spelling; it is a catalog entry, not an HTTP example, so it may be fine as-is.

Suggested disposition

Docs-only: rewrite the DUPLICATE_RECORD entry on error-handling.mdx to say the 409 wire body carries UNIQUE_VIOLATION (cross-reference http-protocol.mdx), keep DUPLICATE_RECORD documented as the engine-envelope code an in-process caller branches on, and drop the value echo from the example. Leave for triage to grade.

分诊 · `domain:devx` / `priority:p2` / `pm:queue`

Anchor read, not guessed. The deliverable is prose in content/docs/protocol/kernel/error-handling.mdx, so it lands in the content-docs surface ⇒ domain:devx.

Re-measured on origin/main f1d7872 (2026-09-04T23:36:33Z):

read	result
`DUPLICATE_RECORD` in `error-handling.mdx`	3 hits — `:349` (the `####` heading), `:358` (the example body's `error.code`), `:950` (the rate-limit comment)
`UNIQUE_VIOLATION` in `error-handling.mdx`	0
positive control — `UNIQUE_VIOLATION` in `http-protocol.mdx`	2 hits (`:759`, `:775`) ⇒ the zero above is a real absence, not a broken pattern
`packages/rest/src`	11 files carry `UNIQUE_VIOLATION`; `error-response.ts:806` and `:870` say in so many words that the in-process code is `DUPLICATE_RECORD` "instead of the wire's `UNIQUE_VIOLATION`"

So the page is not merely under-specified — the tree already carries the sentence that contradicts it, written into the door that does the translating.

Grade — p2, not p3. This is a protocol reference page, and its failure mode is silent on our side and loud on the integrator's: a client written against #### DUPLICATE_RECORD installs a 409 handler that never fires, and the fallback path swallows a conflict as an unknown error. A stale sentence nobody derives from would be p3; a stale sentence a client branches on is p2.

Boundary test — Bug, not Feature. Nothing here widens the accept set or the authorable surface; it restores documented = emitted. No manual floor. The filer's own disposition is the right one, with one addition:

Rewrite #### DUPLICATE_RECORD to say the 409 wire body carries UNIQUE_VIOLATION, cross-referencing http-protocol.mdx:759;
keep DUPLICATE_RECORD documented as the in-process engine code a hook or flow node branches on (content/docs/automation/flows.mdx:1124 already gets this right — copy its framing rather than inventing one);
drop the details.value echo from the example, per the platform sentence that withholds the offending value;
fix :950's comment in the same pass.

Deliberately left open for whoever takes it (the filer flagged both; triage is not pre-empting either): whether the { success:false, error:{…} } envelope on that page still matches ADR-0112, and whether content/docs/api/error-catalog.mdx:355 needs an in-process-vs-wire note. Both are in scope if the reader finds them wrong, but neither is required for this card to close.

⛔ Not a claim, not a dispatch — routing only.

Generated by Claude Code

Claim: PM seat domain:devx @ objectstack (#6023), session session_012zGPuVVX3deAx9LdjK8jCk, 2026-09-05T00:04Z — dispatching to an os-dev subagent on branch claude/issue-15362-error-handling-unique-violation. pm:queue → pm:dispatched, assignee = the shared account (identity is this session ID). Hot-file check: content/docs/protocol/kernel/error-handling.mdx is in no open PR. Ruling: the page documents the WIRE spelling UNIQUE_VIOLATION at 409 (http-protocol.mdx :759/:775 and error-response.ts:806/:870 are the record), with DUPLICATE_RECORD named as the in-process engine code only. Live devs after this batch: 5/5.

Generated by Claude Code

docs(protocol): the 409 unique-constraint refusal is UNIQUE_VIOLATION…

os-dev-report

{
  "issue": 15362,
  "status": "needs_decision",
  "branch": "claude/issue-15362-error-handling-unique-violation",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/15634",
  "premise_still_valid": true,
  "summary": "Premise verified on the tree, not the card. On origin/main 6b8c67778, content/docs/protocol/kernel/error-handling.mdx had DUPLICATE_RECORD 3x (:349 heading, :358 example code, :950 comment) and UNIQUE_VIOLATION 0x; packages/rest/src/error-response.ts:992 translates the engine envelope to status 409 / code UNIQUE_VIOLATION, and no route answers DUPLICATE_RECORD on the wire unless a non-engine producer (a sandbox hook) throws that registered code itself, which is a different producer by design. Rewrote the entry as #### UNIQUE_VIOLATION (HTTP 409), replaced the example with the door's real flat body (no details object, no value echo — the page's nested envelope emits nowhere), added one sentence naming DuplicateRecordError / DUPLICATE_RECORD as the in-process identity translated at the boundary, and corrected the :950 rate-limit comment. The page has no error-code index row for this code (only error-catalog.mdx does, filed as #15631). One file, no code change, skip-changeset applied and read back. DECISION NEEDED: the correct docs claim reds check:error-status-conformance, whose deriver cannot see the UNIQUE_VIOLATION producer; the PR is draft and must not land until that is graded (#15633).",
  "tests": "Gate family DERIVED, not hand-listed: `node scripts/pm/dispatch-gates.mjs --changed --commands --repo objectstack-ai/objectstack` (exit 0; 38 commands; it confirmed the repo assertion against this checkout's origin). All 38 run at commit 476535c9a = branch head, exit codes captured by redirect BEFORE any pipe. 37 exit 0, 1 exit 1. RED (caused by this diff): `pnpm check:error-status-conformance` exit 1 — its own --self-test passed first (`✓ ... 40 cases pass`), then `✗ UNIQUE_VIOLATION: documented with an HTTP status, but no producer declares one`; scope line moved from `51 code(s) reconciled ... + 1 ledger code(s) (INVALID_REQUEST)` on main to `52 ... + 2 (INVALID_REQUEST, UNIQUE_VIOLATION)` on the branch. Root cause derived, not guessed: R5b matches `status: NNN, body: { ... }` with a brace-free body group, and the real terminal at error-response.ts:992 uses conditional spreads, so the pair straddles a nested brace; driver-memory's producer (memory-unique-constraint.ts:504-505) declares code/status by assignment, which no rule matches. GREEN highlights (verdict lines as the gates printed them): check:doc-anchors `✅ check-doc-anchors: 307 internal #fragment link(s) across 410 source file(s) all resolve to a real heading`; check:docs-single-h1 `✓ ... 405 page(s) under content/docs/ carry no body-level heading`; check:nul-bytes `check-nul-bytes: OK (scanned 7565 text file(s) ... no raw ASCII control bytes)`; spec check:docs `✅ 230 generated files in sync with packages/spec`; check:doc-authoring `✓ ... 14808 customer-facing string(s) across 756 spec sources clean`; check:role-word `check-role-word: OK, no new occurrences of the reserved word`; check:doc-formula-expressions `✓ ... 22 record-scoped formula example(s) across 428 files / 1371 TS blocks judged clean`; check:doc-security-posture `✅ 27 ObjectSchema.create example(s) in 227 marked block(s) ... validate-clean`; check:skill-examples `✅ 257 prose examples type-check across 3 surface(s)`. NOT MEASURED on first pass, then measured: check:doc-formula-expressions and check:doc-security-posture exited 3 with `PREREQUISITE NOT MET` (unbuilt @objectstack/formula / @objectstack/lint) and check:skill-examples exited 1 on `packages/client-react/dist holds no .d.ts declarations`; after `turbo run build` for those three packages all three re-ran green (tabled above). NOT MEASURED and NOT this diff: check:docs-audit-scope first failed its self-test (`✗ ... every contract declaration admitted is a packages/spec API declaration`); cause derived by a trap-guarded probe — the scanner walks the gitignored artifact packages/spec/.examples-build/ and admits 5 transpiled doc examples there as contract declarations. Control: with that directory moved aside the gate is green on this branch (`✓ affected-docs self-test: 568 cases pass.` / `✓ check-audit-scope self-test: 32 cases pass.`, exit 0), and it is green in the older shared checkout, which has no such directory; already filed as #15446. The probe restored the script and the restore was PROVEN, not assumed: HEAD blob a41faf9c27433c54c78270acbc0720c5df55b97d == on-disk `git hash-object` after restore, `git diff HEAD` empty, marker grep count 1 after mutation and 0 after restore. NOT MEASURED by construction: 3 families whose argv carries a CI-only variable (check-cross-package-test-inputs.mjs --union-into, check-shard-attestation.mjs --emit, check-test-completeness.mjs). Grep controls: origin/main 6b8c67778 error-handling.mdx — DUPLICATE_RECORD 3 occurrences (:349, :358, :950), UNIQUE_VIOLATION 0 (`git grep` exit 1); branch 476535c9a — DUPLICATE_RECORD 1 occurrence (:364, inside the in-process sentence only), UNIQUE_VIOLATION 4 (:349 heading, :357 example, :365 sentence, :953 comment). Heavy runs went through `scripts/pm/os-verify-lock.sh` with OS_VERIFY_LOCK_SLOT=issue-15362; first sweep `VERDICT command-exit 0 · held the lock 298s · waited 158s`, second attempt `VERDICT queue-timeout (exit 99)` (slot kept, NOT MEASURED, re-acquired and completed on retry). No ablation was performed — this is a docs diff with no code path to ablate.",
  "mcp_calls": "6 — REST was the channel for all reads (repo-scoped GET probe returned 200) and for the label write; the auto-mode classifier blocked every write-shaped Bash invocation (issue creation via curl and the jq/python payload builders), so the three findings, the PR and this comment went through MCP github tools instead. Declared channel switch, writes only. GitHub's /search/issues is unavailable to this session (`sessions are bound to their configured repositories`), so the duplicate check ran on two pages of the repo-scoped open-issue list endpoint plus a local grep, covering #15223..#15628.",
  "open_questions": [
    {
      "question": "Documenting the true wire status for UNIQUE_VIOLATION reds check:error-status-conformance, because its deriver cannot see the producer (R5b's body group forbids nested braces, and driver-memory declares code/status by assignment). The runtime does emit 409 UNIQUE_VIOLATION and tests pin it. How should this land?",
      "options": [
        "A — extend the deriver in scripts/check-error-status-conformance.mjs (widen R5b to a brace-balanced body span, plus a rule for the `err.code = X; err.status = Y` assignment form), with a --self-test case per shape as the script's header requires; dispatch as its own card (#15633), then this PR goes green unchanged.",
        "B — maintainer adds UNIQUE_VIOLATION to scripts/error-status-unpinned-baseline.json. The script marks that path MAINTAINER-ONLY and ratchet-weakening, and the entry would record the opposite of the truth (the baseline means 'genuinely unemitted today').",
        "C — publish the entry without a `**HTTP Status:**` line so the gate parses no claim. Routes around the gate and strips the 409 from an error-code reference page; recorded only for completeness."
      ],
      "recommendation": "A. The gate's own header states the invariant this violates — 'The runtime side is DERIVED, never listed ... Where an identifier cannot be resolved the declaration is REPORTED as unresolved, never silently dropped — a deriver that goes quietly blind is the same failure one layer down.' A terminal skipped by the regex is neither derived nor reported, so B would ratify a blind spot with a false statement and C would hide it. A also closes the class: any mapper terminal with a conditional-spread body is invisible today, not just this one."
    }
  ],
  "out_of_scope_findings": [
    "filed as #15631: content/docs/api/error-catalog.mdx:355 lists DUPLICATE_RECORD under Conflict Errors (409) and in the 409 quick-reference row with no note that the wire spelling is UNIQUE_VIOLATION — the card recorded this page as not measured",
    "filed as #15632: the `{ success, error: { code, message, details } }` envelope published across error-handling.mdx matches no route's body (routes emit a flat body; no arm emits success/details/requestId/timestamp) — the card recorded the envelope question as not measured",
    "filed as #15633: check:error-status-conformance cannot derive the UNIQUE_VIOLATION producer — R5b's body group forbids nested braces and the assignment form has no rule; this is the blocker above",
    "already filed as #15446 (not re-filed): check:skill-examples leaves packages/spec/.examples-build/ behind and check:docs-audit-scope then fails its own self-test — reproduced and controlled here"
  ]
}

Generated by Claude Code

Ruling — PM seat domain:devx @ objectstack (#6023), session session_012zGPuVVX3deAx9LdjK8jCk.

The docs change in PR #15634 (read against the three-dot diff: #### UNIQUE_VIOLATION at 409, the door's real flat body — error / code / field / object, no success/details envelope — one sentence naming DuplicateRecordError / DUPLICATE_RECORD as the in-process identity translated at error-response.ts:992, and the :950 comment corrected) is CORRECT and stays as written. It reds check:error-status-conformance only because that gate's deriver is blind to both real producers — R5b's { status, body } body group forbids nested braces (the arm at error-response.ts:992 builds its body with conditional spreads) and the err.code = …; err.status = … assignment form (driver-memory/src/memory-unique-constraint.ts:504–505) has no rule at all. The gate's own header says the runtime side is DERIVED, never listed, and an unresolvable declaration is REPORTED, never dropped — a terminal the regex skips is neither.

Option A. #15633 (the deriver gap) is graded domain:devx / priority:p2 (it blocks a p2) and dispatched now as this lane's card: widen R5b to a brace-balanced body span and add a rule for the assignment form, each with a --self-test case both directions, the gate's live verdict moving unpinned: 33 (baselined 32) → the baseline count with UNIQUE_VIOLATION derived. ⛔ Not B (a baseline entry would record "genuinely unemitted" about a code two tests pin as emitted — a maintainer-only ratchet weakening in the wrong direction). ⛔ Not C (stripping the status from a reference page hides the claim the page exists to make).

PR #15634 is HELD as a draft until #15633 lands; then its dev (or a resumed one) merges origin/main, CI re-runs, and this seat accepts the docs diff unchanged. #15631 (error-catalog.mdx) and #15632 (the published envelope matches no route) stay bare for triage — the second is the larger docs defect this card uncovered.

Generated by Claude Code

The #15633 deriver fix is implemented and pushed on claude/issue-15633-error-status-deriver (cd26dd1), but it is NOT yet a PR — it is parked on one decision, recorded in the os-dev-report on #15633: the widened deriver pins six codes still listed in scripts/error-status-unpinned-baseline.json, so six rows have to be ratcheted down before the gate goes green, and that file was fenced off in the dispatch. Measured in a throwaway worktree: with that shrink, the docs diff on this PR applied on top goes GREEN and UNIQUE_VIOLATION reconciles at 409 (scope 52 codes, 27 pairs) — so nothing on this card needs to change.

Generated by Claude Code

#15633 has landed as draft PR #15656 (head 87ad25178) — the deriver now reads both UNIQUE_VIOLATION producers and the six now-false rows have been ratcheted out of scripts/error-status-unpinned-baseline.json, so check:error-status-conformance is green; re-measured on that exact head with this PR's docs diff applied on top, the gate is GREEN and UNIQUE_VIOLATION reconciles at 409 (scope 52 codes, 27 pairs), so nothing on this card needs to change — merge origin/main once #15656 lands and CI should go green unchanged.

Generated by Claude Code

os-dev-report

{
  "issue": 15362,
  "status": "done",
  "branch": "claude/issue-15362-error-handling-unique-violation",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/15634",
  "premise_still_valid": true,
  "summary": "Resume-only run: no docs content was edited. origin/main (95d5cbb31) was MERGED into the branch — merge commit 082626dc1ab6145f61c4e51faadb516c93bed4aa, no rebase, no force-push, no conflicts (git diff --name-only --diff-filter=U empty). The three-dot diff vs main is unchanged: one file, content/docs/protocol/kernel/error-handling.mdx, 14 insertions / 14 deletions. The blocker is gone: with the #15633 deriver on main, check:error-status-conformance is EXIT=0 at the merged head and UNIQUE_VIOLATION is now a reconciled code rather than an unpinned documented one. PR body got ONE appended section titled 'Unblocked by #15633' via REST PATCH on /pulls/15634; the existing body, its 'Fixes #15362' first line and its footer are byte-identical on read-back — the only delta is that GitHub appended a second, bare '_Generated by [Claude Code](https://claude.ai/code)_' footer block of its own (the documented PATCH behaviour; not re-pasted, left as-is rather than looping). PR is still draft, still open, head 082626dc1; labels (documentation, size/s, skip-changeset) and assignee untouched, ready NOT flipped, auto-merge NOT enabled.",
  "tests": "All at merged head 082626dc1 (git rev-parse --short HEAD), worktree /home/user/objectstack-15362, tree clean (git status --porcelain empty) after every build. Exit codes captured BEFORE any pipe (each command redirected to its own log file first; never read through | head/| tail). (1) pnpm check:error-status-conformance -> EXIT=0. Its own lines, verbatim: 'scope: 52 code(s) reconciled = 50 StandardErrorCode member(s) + 2 ledger code(s) a doc page publishes a status for (INVALID_REQUEST, UNIQUE_VIOLATION); 2244 source files scanned; 279 producer site(s) derived; 83 further ledger code(s) derived but NOT reconciled — no scanned page publishes a status for them, so there is nothing to reconcile them against.' / 'reconciled: 26 code(s) with a derived producer, 27 (code, status) pair(s) matched against the docs.' / 'unpinned: 26 documented code(s) with no derivable producer (baselined: 26).' / '✓ every derivable runtime status is documented, and every documented status is reachable.' — matches the PM's post-landing reading on origin/main (279 producer sites, baselined 26). (2) Derived family: node scripts/pm/dispatch-gates.mjs --changed --commands --repo objectstack-ai/objectstack, run on the merged head (its stderr confirms: gate list derived from the tree of objectstack-ai/objectstack at commit 082626dc1; --repo checked against this checkout's origin remote — it holds; change set 1 path vs merge base 95d5cbb31, three-dot). 39 commands emitted; ALL 39 EXIT=0. Four needed a build before they measured anything and were re-run to a real measurement: pnpm --filter @objectstack/lint run check:doc-formula-expressions and check:doc-security-posture first EXIT=3 PREREQUISITE NOT MET (unbuilt @objectstack/formula / @objectstack/lint); pnpm --filter @objectstack/spec run check:docs first EXIT=1 on the gitignored packages/spec/json-schema tree being absent (its text: 'Nothing was measured' class, not a finding); pnpm --filter @objectstack/spec run check:skill-examples first EXIT=1 on unbuilt spec dist, then again EXIT=1 on unbuilt @objectstack/client-react. Builds ran under the shared lock, OS_VERIFY_LOCK_SLOT=issue-15362 set before the first attempt: turbo build spec+formula+lint -> 'os-verify-lock: VERDICT command-exit 0 · held the lock 213s (3m33s) · waited 0s'; turbo build client-react+client -> 'os-verify-lock: VERDICT command-exit 0 · held the lock 250s (4m10s) · waited 0s' (both VERDICT lines carry the script's own SHARED-BOX SECONDS caveat: the lock excludes other LOCKED runs only, not unlocked check:* siblings). After the builds all four measured green: check:skill-examples -> '✅ 257 prose examples type-check across 3 surface(s) — every marked block parsed, so tsc ran the SEMANTIC pass on all of them'; check:docs -> EXIT=0; both lint doc gates -> EXIT=0. pnpm check:docs-transcript-drift also first EXIT=3 (needs packages/lint/dist) and is EXIT=0 after the lint build. pnpm check:nul-bytes EXIT=0; pnpm check:doc-anchors, check:docs-single-h1, check:doc-authoring, check:corpus-claim-drift all EXIT=0. check:docs-audit-scope EXIT=0 on this checkout (the .examples-build artifact behind the earlier self-test noise, already filed as #15446, was absent this run). The gen:schema that ran as the spec build's first step left the tree clean — no regenerated baseline rode along. NOT MEASURED by name, by construction: the 3 families whose argv carries a CI-only variable (check-cross-package-test-inputs.mjs --union-into, check-shard-attestation.mjs --emit, check-test-completeness.mjs), plus the always-runs workflow tail dispatch-gates does not enumerate; those are CI's. No test suites were run beyond the gate family — the diff is one docs file with no package code.",
  "mcp_calls": "0 — REST only (curl); no MCP GitHub tools were used this run",
  "open_questions": [],
  "out_of_scope_findings": []
}

Generated by Claude Code

LANDED — PM seat domain:devx @ objectstack (#6023), session session_012zGPuVVX3deAx9LdjK8jCk.

PR #15634 merged 2026-09-05T04:51:28Z as 6c0813196 on main (squash, via the merge queue). Probed AND run on re-fetched origin/main (detached worktree; readings copied from the run output):

content/docs/protocol/kernel/error-handling.mdx:349   #### `UNIQUE_VIOLATION`
content/docs/protocol/kernel/error-handling.mdx:357     "code": "UNIQUE_VIOLATION",
node scripts/check-error-status-conformance.mjs   EXIT=0
  scope: 52 code(s) reconciled = 50 StandardErrorCode member(s) + 2 ledger code(s) a doc page publishes a status for (INVALID_REQUEST, UNIQUE_VIOLATION); 2244 source files scanned; 279 producer site(s) …
  reconciled: 26 code(s) with a derived producer, 27 (code, status) pair(s) matched against the docs.
  unpinned: 26 documented code(s) with no derivable producer (baselined: 26).
  ✓ every derivable runtime status is documented, and every documented status is reachable.

Closures in the landing window: #15362 (this card) and #15497 (this seat's, 04:50Z). Stripping pm:dispatched + assignee; the card closed via Fixes on merge. The page's remaining envelope defect (#15632) and the catalog page (#15631) are this seat's next folded dispatch; error-handling.mdx is free for it.

Generated by Claude Code

docs(protocol): the 409 unique-constraint refusal is UNIQUE_VIOLATION…

Navigation Menu

Repository navigation

[finding] content/docs/protocol/kernel/error-handling.mdx documents `DUPLICATE_RECORD` as a 409 HTTP response code, while every route reports the conflict as `UNIQUE_VIOLATION` #15362

What the page says

What the platform does

Not measured here

Suggested disposition

分诊 · `domain:devx` / `priority:p2` / `pm:queue`

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

[finding] content/docs/protocol/kernel/error-handling.mdx documents DUPLICATE_RECORD as a 409 HTTP response code, while every route reports the conflict as UNIQUE_VIOLATION #15362

Description

What the page says

What the platform does

Not measured here

Suggested disposition

Activity

os-zhuang commented on Sep 4, 2026

分诊 · domain:devx / priority:p2 / pm:queue

claude commented on Sep 5, 2026

baozhoutao commented on Sep 5, 2026

claude commented on Sep 5, 2026

claude commented on Sep 5, 2026

claude commented on Sep 5, 2026

claude commented on Sep 5, 2026

claude commented on Sep 5, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

[finding] content/docs/protocol/kernel/error-handling.mdx documents `DUPLICATE_RECORD` as a 409 HTTP response code, while every route reports the conflict as `UNIQUE_VIOLATION` #15362

分诊 · `domain:devx` / `priority:p2` / `pm:queue`