test(layout): move providers and codex-integration into tests/<domain>/ (#3497)

.github/scripts/pr-hygiene.test.cjs

@@ -225,7 +225,7 @@ describe("collectDeterministicHygieneFailures", () => {
     const failures = collectDeterministicHygieneFailures({
       files: [
         { filename: "src/codex/auth-api.ts", patch: "+change" },
-        { filename: "tests/codex-auth-api.test.ts", patch: "+test" },
+        { filename: "tests/codex-integration/codex-auth-api.test.ts", patch: "+test" },
       ],
       authorHasPushPermission: true,
     });

docs/superpowers/plans/2026-07-26-oauth-reliability-integrity.md

@@ -209,7 +209,7 @@ EOF
 
 **Files:**
 - Modify: `src/oauth/index.ts` (`refreshAndPersistAccessToken` generic branch ~352–400)
-- Test: `tests/oauth/oauth-refresh-generic-lock.test.ts` (new; mirror patterns from `tests/xai-refresh-lock.test.ts` / `tests/oauth/oauth-refresh.test.ts`)
+- Test: `tests/oauth/oauth-refresh-generic-lock.test.ts` (new; mirror patterns from `tests/providers/xai/xai-refresh-lock.test.ts` / `tests/oauth/oauth-refresh.test.ts`)
 
 **Interfaces:**
 - Consumes: `createOAuthRefreshIntentLock`, `mergeAccountCredential`, `credentialGeneration`, `markAccountNeedsReauthIfGeneration`, `getAccountCredential`, `logOAuthEvent`
@@ -308,7 +308,7 @@ Also log `"OAuth refresh joined existing operation"` when `tokenRefreshes.get(ke
 Run:
 
 ```bash
-bun test tests/oauth/oauth-refresh-generic-lock.test.ts tests/oauth/oauth-refresh.test.ts tests/xai-refresh-lock.test.ts
+bun test tests/oauth/oauth-refresh-generic-lock.test.ts tests/oauth/oauth-refresh.test.ts tests/providers/xai/xai-refresh-lock.test.ts
 ```
 
 Expected: PASS (no regressions on xAI/Anthropic)
@@ -420,7 +420,7 @@ Set `action` strings:
 
 - [ ] **Step 4: Run test to verify it passes**
 
-Run: `bun test tests/oauth/oauth-health.test.ts tests/codex-routing.test.ts`
+Run: `bun test tests/oauth/oauth-health.test.ts tests/codex-integration/codex-routing.test.ts`
 
 Expected: PASS
 
@@ -502,7 +502,7 @@ EOF
 
 **Files:**
 - Modify: `src/cli/doctor.ts`
-- Test: `tests/service/doctor-oauth.test.ts` (or extend `tests/doctor.test.ts`)
+- Test: `tests/service/doctor-oauth.test.ts` (or extend `tests/codex-integration/doctor.test.ts`)
 
 **Interfaces:**
 - Consumes: `collectOAuthHealthEntries`, auth store writability checks, refresh lock path helpers if exported
@@ -529,7 +529,7 @@ Add `collectOAuthDoctorChecks(): Array<{ level: "OK" | "WARN"; message: string }
 
 - [ ] **Step 4: Run tests**
 
-Run: `bun test tests/service/doctor-oauth.test.ts tests/doctor.test.ts`
+Run: `bun test tests/service/doctor-oauth.test.ts tests/codex-integration/doctor.test.ts`
 
 Expected: PASS
 
@@ -607,9 +607,9 @@ EOF
 ### Task 8: Codex metadata integrity regressions + 401 replay invariants
 
 **Files:**
-- Test: `tests/codex-metadata-integrity.test.ts` (new)
+- Test: `tests/codex-integration/codex-metadata-integrity.test.ts` (new)
 - Modify only if a real gap is found: `src/codex/auth-context.ts`, `src/adapters/openai-responses.ts`
-- Confirm existing: `tests/server-xai-oauth-401-replay.test.ts`, `tests/server-kiro-oauth-401-replay.test.ts`, `tests/codex-routing.test.ts` (policy A)
+- Confirm existing: `tests/server-xai-oauth-401-replay.test.ts`, `tests/server-kiro-oauth-401-replay.test.ts`, `tests/codex-integration/codex-routing.test.ts` (policy A)
 
 **Interfaces:**
 - Consumes: `headersForCodexAuthContext`, `FORWARD_HEADERS`
@@ -644,7 +644,7 @@ test("preserves genuine originator", () => {
 
 - [ ] **Step 2: Run tests**
 
-Run: `bun test tests/codex-metadata-integrity.test.ts`
+Run: `bun test tests/codex-integration/codex-metadata-integrity.test.ts`
 
 Expected: FAIL only if implementation gap exists; if PASS immediately, keep tests as regressions and skip code changes.
 
@@ -655,15 +655,15 @@ If fabrication or account-id mismatch is found, fix the minimal header path. Do
 - [ ] **Step 4: Run related suite**
 
 ```bash
-bun test tests/codex-metadata-integrity.test.ts tests/codex-auth-context.test.ts tests/codex-routing.test.ts tests/session-affinity.test.ts tests/server-xai-oauth-401-replay.test.ts tests/server-kiro-oauth-401-replay.test.ts
+bun test tests/codex-integration/codex-metadata-integrity.test.ts tests/codex-integration/codex-auth-context.test.ts tests/codex-integration/codex-routing.test.ts tests/session-affinity.test.ts tests/server-xai-oauth-401-replay.test.ts tests/server-kiro-oauth-401-replay.test.ts
 ```
 
 Expected: PASS
 
 - [ ] **Step 5: Commit**
 
 ```bash
-git add tests/codex-metadata-integrity.test.ts src/codex/auth-context.ts src/adapters/openai-responses.ts
+git add tests/codex-integration/codex-metadata-integrity.test.ts src/codex/auth-context.ts src/adapters/openai-responses.ts
 git commit -m "$(cat <<'EOF'
 test(codex): lock metadata pass-through and non-fabrication
 
@@ -713,8 +713,8 @@ EOF
 - [ ] **Step 1: Run verification commands**
 
 ```bash
-bun test tests/lib/privacy-mask-account.test.ts tests/oauth/oauth-log.test.ts tests/oauth/oauth-refresh-generic-lock.test.ts tests/oauth/oauth-health.test.ts tests/cli/cli-status-oauth-health.test.ts tests/service/doctor-oauth.test.ts tests/oauth/oauth-accounts-api.test.ts tests/codex-metadata-integrity.test.ts
-bun test tests/oauth/oauth-refresh.test.ts tests/xai-refresh-lock.test.ts tests/codex-routing.test.ts tests/session-affinity.test.ts tests/codex-auth-context.test.ts
+bun test tests/lib/privacy-mask-account.test.ts tests/oauth/oauth-log.test.ts tests/oauth/oauth-refresh-generic-lock.test.ts tests/oauth/oauth-health.test.ts tests/cli/cli-status-oauth-health.test.ts tests/service/doctor-oauth.test.ts tests/oauth/oauth-accounts-api.test.ts tests/codex-integration/codex-metadata-integrity.test.ts
+bun test tests/oauth/oauth-refresh.test.ts tests/providers/xai/xai-refresh-lock.test.ts tests/codex-integration/codex-routing.test.ts tests/session-affinity.test.ts tests/codex-integration/codex-auth-context.test.ts
 bun run test
 bun run typecheck
 bun run lint:gui

docs/superpowers/plans/2026-08-06-deepseek-responses-streaming-terminal-repair.md

@@ -27,7 +27,7 @@
 - `src/server/responses-terminal-repair.ts` — owns SSE lifecycle tracking, bounded retained state, grace scheduling, and synthetic terminal creation.
 - `src/server/responses/core.ts` — activates the repair before existing transport-specific relay branches.
 - `tests/responses/responses-terminal-repair.test.ts` — unit state-machine and stream-race coverage.
-- `tests/deepseek-inbound-wire.test.ts` — end-to-end official DeepSeek wire and HTTP activation.
+- `tests/providers/deepseek-inbound-wire.test.ts` — end-to-end official DeepSeek wire and HTTP activation.
 - `tests/responses/ws-endpoint.test.ts` — WebSocket event parity for real and repaired terminals.
 - `structure/04_transports-and-sidecars.md` — architectural contract for the provider-scoped streaming repair.
 - `docs/superpowers/specs/2026-08-06-deepseek-responses-streaming-terminal-repair-design.md` — approved design authority; implementation must remain consistent with it.
@@ -40,8 +40,8 @@
 - Modify: `src/providers/registry.ts:150-170`
 - Modify: `src/providers/registry.ts:1143-1162`
 - Modify: `src/providers/registry.ts:1834-1842`
-- Modify: `tests/deepseek-inbound-wire.test.ts:1-175`
-- Modify: `tests/deepseek-responses-item-id-repair.test.ts`
+- Modify: `tests/providers/deepseek-inbound-wire.test.ts:1-175`
+- Modify: `tests/providers/deepseek-responses-item-id-repair.test.ts`
 
 **Interfaces:**
 - Produces: `ResponsesTerminalRepairPolicy`
@@ -74,7 +74,7 @@ test("Codex HTTP and WebSocket turns keep DeepSeek streaming upstream", async ()
 Delete or rewrite the tests whose asserted contract is specifically
 `stream:false`/bounded JSON for built-in DeepSeek. Keep the bounded-JSON helper
 coverage that is provider-neutral. In
-`tests/deepseek-responses-item-id-repair.test.ts`, retain the pure
+`tests/providers/deepseek-responses-item-id-repair.test.ts`, retain the pure
 `repairResponsesJsonItemIds()` unit test but remove the built-in-DeepSeek HTTP
 activation assertion; Task 4 replaces it with streaming repair composition.
 
@@ -83,7 +83,7 @@ activation assertion; Task 4 replaces it with streaming repair composition.
 Run:
 
 ```bash
-bun test tests/deepseek-inbound-wire.test.ts
+bun test tests/providers/deepseek-inbound-wire.test.ts
 ```
 
 Expected: compile/test failure because `providerModelResponsesTerminalRepair`
@@ -134,7 +134,7 @@ export function providerModelResponsesTerminalRepair(
 Run:
 
 ```bash
-bun test tests/deepseek-inbound-wire.test.ts tests/deepseek-responses-item-id-repair.test.ts tests/provider-registry-parity.test.ts
+bun test tests/providers/deepseek-inbound-wire.test.ts tests/providers/deepseek-responses-item-id-repair.test.ts tests/providers/provider-registry-parity.test.ts
 ```
 
 Expected: all selected tests pass; captured HTTP and WebSocket request bodies
@@ -143,7 +143,7 @@ carry `stream:true`.
 - [ ] **Step 5: Commit**
 
 ```bash
-git add src/providers/registry.ts tests/deepseek-inbound-wire.test.ts tests/deepseek-responses-item-id-repair.test.ts
+git add src/providers/registry.ts tests/providers/deepseek-inbound-wire.test.ts tests/providers/deepseek-responses-item-id-repair.test.ts
 git commit -m "fix(deepseek): restore Responses upstream streaming"
 ```
 
@@ -386,8 +386,8 @@ git commit -m "fix(responses): fail closed on unsafe terminal repair"
 **Files:**
 - Modify: `src/server/responses/core.ts:95-115`
 - Modify: `src/server/responses/core.ts:2030-2230`
-- Modify: `tests/deepseek-inbound-wire.test.ts:115-330`
-- Modify: `tests/deepseek-responses-item-id-repair.test.ts`
+- Modify: `tests/providers/deepseek-inbound-wire.test.ts:115-330`
+- Modify: `tests/providers/deepseek-responses-item-id-repair.test.ts`
 - Modify: `tests/responses/ws-endpoint.test.ts`
 
 **Interfaces:**
@@ -432,7 +432,7 @@ WebSocket handling. Assert the client receives progressive delta frames, one
 Run:
 
 ```bash
-bun test tests/deepseek-inbound-wire.test.ts tests/responses/ws-endpoint.test.ts
+bun test tests/providers/deepseek-inbound-wire.test.ts tests/responses/ws-endpoint.test.ts
 ```
 
 Expected: requests now carry `stream:true` from Task 1 but no provider-scoped
@@ -482,7 +482,7 @@ streaming resolver still returns `false`.
 Run:
 
 ```bash
-bun test tests/deepseek-inbound-wire.test.ts tests/responses/ws-endpoint.test.ts tests/responses/responses-state.test.ts tests/deepseek-responses-item-id-repair.test.ts tests/deepseek-reasoning-replay.test.ts
+bun test tests/providers/deepseek-inbound-wire.test.ts tests/responses/ws-endpoint.test.ts tests/responses/responses-state.test.ts tests/providers/deepseek-responses-item-id-repair.test.ts tests/providers/deepseek-reasoning-replay.test.ts
 ```
 
 Expected: progressive output precedes terminal, both transports close once,
@@ -491,7 +491,7 @@ item ids are stable, and continuation state retains reasoning/function output.
 - [ ] **Step 6: Commit**
 
 ```bash
-git add src/server/responses/core.ts tests/deepseek-inbound-wire.test.ts tests/deepseek-responses-item-id-repair.test.ts tests/responses/ws-endpoint.test.ts
+git add src/server/responses/core.ts tests/providers/deepseek-inbound-wire.test.ts tests/providers/deepseek-responses-item-id-repair.test.ts tests/responses/ws-endpoint.test.ts
 git commit -m "fix(deepseek): repair terminal-less Responses streams"
 ```
 
@@ -560,7 +560,7 @@ git commit -m "docs(deepseek): describe streaming terminal repair"
 Run:
 
 ```bash
-bun test tests/responses/responses-terminal-repair.test.ts tests/deepseek-inbound-wire.test.ts tests/responses/ws-endpoint.test.ts tests/responses/sse-failed-tail.test.ts tests/relay-eager.test.ts tests/responses/responses-item-id-repair.test.ts tests/deepseek-responses-item-id-repair.test.ts tests/deepseek-reasoning-replay.test.ts tests/responses/responses-state.test.ts
+bun test tests/responses/responses-terminal-repair.test.ts tests/providers/deepseek-inbound-wire.test.ts tests/responses/ws-endpoint.test.ts tests/responses/sse-failed-tail.test.ts tests/relay-eager.test.ts tests/responses/responses-item-id-repair.test.ts tests/providers/deepseek-responses-item-id-repair.test.ts tests/providers/deepseek-reasoning-replay.test.ts tests/responses/responses-state.test.ts
 ```
 
 Expected: 0 failures.

docs/superpowers/plans/2026-08-09-routed-computer-use-browser.md

@@ -13,9 +13,9 @@
 ## Task 1: Lock the routed catalog contract with failing tests
 
 **Files:**
-- Modify: `tests/codex-catalog.test.ts:280-300`
-- Modify: `tests/codex-catalog.test.ts:1011-1045`
-- Modify: `tests/codex-catalog.test.ts:1230-1255`
+- Modify: `tests/codex-integration/codex-catalog.test.ts:280-300`
+- Modify: `tests/codex-integration/codex-catalog.test.ts:1011-1045`
+- Modify: `tests/codex-integration/codex-catalog.test.ts:1230-1255`
 
 **Step 1: Change the combo assertion to require the explicit routed mode**
 
@@ -68,15 +68,15 @@ expect(rows.find(row => row.slug === "team/gpt-5.5")?.tool_mode).toBe("code");
 Run:
 
 ```bash
-bun test tests/codex-catalog.test.ts
+bun test tests/codex-integration/codex-catalog.test.ts
 ```
 
 Expected: routed/template/combo/fallback assertions fail because current production code deletes or omits `tool_mode`; native preservation assertions remain green.
 
 **Step 6: Commit the red tests only**
 
 ```bash
-git add tests/codex-catalog.test.ts
+git add tests/codex-integration/codex-catalog.test.ts
 git commit -m "test(codex): require code mode for routed models"
 ```
 
@@ -86,7 +86,7 @@ git commit -m "test(codex): require code mode for routed models"
 - Modify: `src/codex/catalog/parsing.ts:341-375`
 - Modify: `src/codex/catalog/sync.ts:33`
 - Modify: `src/codex/catalog/sync.ts:276-310`
-- Test: `tests/codex-catalog.test.ts`
+- Test: `tests/codex-integration/codex-catalog.test.ts`
 
 **Step 1: Add a named compatibility helper**
 
@@ -128,7 +128,7 @@ Do not call the full routed normalizer from the fallback branch; that would broa
 **Step 4: Run the focused tests and confirm GREEN**
 
 ```bash
-bun test tests/codex-catalog.test.ts
+bun test tests/codex-integration/codex-catalog.test.ts
 ```
 
 Expected: all catalog tests pass, including native/account preservation.
@@ -151,8 +151,8 @@ git commit -m "fix(codex): enable code mode for routed models"
 ## Task 3: Prove on-disk catalog synchronization preserves the policy
 
 **Files:**
-- Modify: `tests/codex-catalog-sync-hardening.test.ts`
-- Test: `tests/codex-catalog-sync-hardening.test.ts`
+- Modify: `tests/codex-integration/codex-catalog-sync-hardening.test.ts`
+- Test: `tests/codex-integration/codex-catalog-sync-hardening.test.ts`
 
 **Step 1: Add a real sync fixture assertion**
 
@@ -169,15 +169,15 @@ Also assert a generated native account-qualified row retains `"code"` if the fix
 **Step 2: Run the sync test**
 
 ```bash
-bun test tests/codex-catalog-sync-hardening.test.ts
+bun test tests/codex-integration/codex-catalog-sync-hardening.test.ts
 ```
 
 Expected: pass and prove the serialized catalog, not merely the in-memory builder.
 
 **Step 3: Commit the persistence test**
 
 ```bash
-git add tests/codex-catalog-sync-hardening.test.ts
+git add tests/codex-integration/codex-catalog-sync-hardening.test.ts
 git commit -m "test(codex): persist routed code mode policy"
 ```
 
@@ -229,7 +229,7 @@ git commit -m "docs(codex): explain routed local tool access"
 Run:
 
 ```bash
-bun test tests/responses/responses-parser.test.ts tests/adapters/bridge.test.ts tests/multi-agent-compat.test.ts
+bun test tests/responses/responses-parser.test.ts tests/adapters/bridge.test.ts tests/codex-integration/multi-agent-compat.test.ts
 ```
 
 Expected: custom declaration, streaming `custom_tool_call`, exact freeform input, and output replay remain green. Add no production protocol change unless one of these tests proves a real gap.
@@ -239,7 +239,7 @@ Expected: custom declaration, streaming `custom_tool_call`, exact freeform input
 Run:
 
 ```bash
-bun test tests/responses/responses-terminal-repair.test.ts tests/deepseek-inbound-wire.test.ts tests/deepseek-responses-item-id-repair.test.ts tests/responses/passthrough-abort.test.ts
+bun test tests/responses/responses-terminal-repair.test.ts tests/providers/deepseek-inbound-wire.test.ts tests/providers/deepseek-responses-item-id-repair.test.ts tests/responses/passthrough-abort.test.ts
 ```
 
 Expected: progressive deltas, strict terminal repair, item IDs, and cancellation pass.
@@ -249,7 +249,7 @@ Expected: progressive deltas, strict terminal repair, item IDs, and cancellation
 Run:
 
 ```bash
-bun test tests/vision/vision-sidecar-e2e.test.ts tests/vision/vision-anthropic.test.ts tests/vision/vision-cache.test.ts tests/vision/vision-fail-closed.test.ts tests/catalog-vision-sidecar-modalities.test.ts tests/responses/openai-responses-passthrough.test.ts
+bun test tests/vision/vision-sidecar-e2e.test.ts tests/vision/vision-anthropic.test.ts tests/vision/vision-cache.test.ts tests/vision/vision-fail-closed.test.ts tests/codex-integration/catalog-vision-sidecar-modalities.test.ts tests/responses/openai-responses-passthrough.test.ts
 ```
 
 Expected: captions replace raw image parts in passthrough bodies, empty references do not consume captions, and partial/failure paths omit pixels safely.

docs/superpowers/plans/2026-08-13-cl10-deep-review-hardening.md

@@ -175,7 +175,7 @@
 ### Task 12: Move catalog timeout to the flaky test only
 
 **Files:**
-- Modify: `tests/codex-catalog-sync-hardening.test.ts`
+- Modify: `tests/codex-integration/codex-catalog-sync-hardening.test.ts`
 - Modify: `scripts/ci/run-bun-test-batches.sh`
 
 **Interfaces:**