RI-7682 Consume auto-generated apiClient types in the UI

[KrumTy]

What

Phase 2 of RI-7682 — replace every UI import from apiSrc/* / src/* (BE) with the auto-generated apiClient types produced by the postinstall script that landed in #5863. Flip the no-restricted-imports lint rule for those patterns from warn to error so new BE -> FE coupling fails CI.

Where the OpenAPI surface doesn't cover the type — socket.io payloads (IMonitorData, PubSubMessage, SubscriptionDto), the IBulkActionOverview family that the bulk-import controller declares as type: Object in swagger, and the Azure OAuth callback values that ship via localStorage rather than HTTP — define FE-local equivalents in their existing slices/interfaces/* files (or constants/azure.ts for the Azure pair). Two stale imports that pointed at BE paths that no longer exist got fixed in passing: AddSentinelMasterResponse → CreateSentinelDatabaseResponse, GetCloudAccountShortInfoResponse → CloudAccountInfo.

Three of the FE-local enum mirrors are on the REST surface and were straightforward to promote — ListElementDestination, NodeRole, HealthStatus now come from apiClient. To make that work, added enumName: 'X' on those 4 @ApiProperty({ enum }) decorator sites (Nest emits the enum as a components/schemas/X reference instead of inlining the union) and configured @hey-api/typescript to emit enums: 'typescript' with case: 'PascalCase', so call sites keep reading NodeRole.Primary rather than the default SCREAMING_SNAKE_CASE.

A follow-up PR will continue the same pattern for the remaining ~30 @ApiProperty({ enum }) decorators that still lack enumName, plus tighten optionality (@ApiProperty → @ApiPropertyOptional where the field is TS-optional) and add @ApiProperty for runtime-present fields that the spec currently omits.

Testing

• grep confirms zero remaining apiSrc/* / src/* imports under redisinsight/ui. A probe import errors out, confirming the guard at error level.
• yarn lint passes.
• No CI gate runs tsc on the UI (verified — lint.yml runs eslint only; tests use babel-jest which strips types; the type-check:ui script is never invoked from a workflow). Build/test behavior is unchanged.
• tsc -p tsconfig.json for context: main has 2498 errors, this branch has 2160 (−338 net). Of those, 795 main-only errors went away, 457 are new. The 457 are real BE↔FE shape-drift findings now visible after the type swap (e.g. RdiInstance.lastConnection is required in swagger but optional in BE class; CommandExecutionResult.sizeLimitExceeded similar; GetUserAgreementsResponse is missing fields that exist at runtime; factories use faker.date.past() where the wire format is string). None of these is a functional regression — type-only — and the follow-up PR addresses them at the source.

---

Refs RI-7682

---

Note

Medium Risk
Medium risk because it touches a large number of UI files and type definitions and tightens linting to fail CI, but changes are largely type/import rewiring with minimal runtime behavior impact (aside from new CI generation step).

Overview
Migrates the UI off backend-internal type imports. Replaces apiSrc/*/src/* imports across the UI with the generated apiClient types, and adds FE-local equivalents where the API surface doesn’t cover the shapes (e.g. bulkActions, monitor, pubsub, and Azure OAuth callback constants).

Hardens the boundary between BE and FE. Updates ESLint no-restricted-imports for UI files to error on apiSrc/* and src/* imports, adjusts OpenAPI client generation to emit named TS enums (PascalCase) and updates a few BE Swagger decorators with enumName so enums like ListElementDestination, NodeRole, and HealthStatus can be consumed from apiClient.

CI reliability update. Extends the shared GitHub Action to generate redisinsight/api-client/ when caches restore node_modules and postinstall doesn’t run, preventing UI builds from failing to resolve apiClient.

^{Reviewed by Cursor Bugbot for commit 6e6588a. Bugbot is set up for automated code reviews on this repo. Configure here.}

[jit-ci]

🛡️ Jit Security Scan Results

✅ No security findings were detected in this PR

---

^{Security scan by Jit}

[github-actions]

Code Coverage - Backend unit tests

St.❔	Category	Percentage	Covered / Total
🟢	Statements	92.57%	15404/16641
🟡	Branches	74.48%	4813/6462
🟢	Functions	86.54%	2391/2763
🟢	Lines	92.39%	14720/15933

Test suite run success

3357 tests passing in 305 suites.

Report generated by 🧪jest coverage report action from 6e6588a

[github-actions]

Code Coverage - Integration Tests

Status	Category	Percentage	Covered / Total
🟡	Statements	79.81%	17433/21843
🟡	Branches	62.26%	7970/12801
🟡	Functions	68.55%	2418/3527
🟡	Lines	79.37%	16390/20649

[github-actions]

Code Coverage - Frontend unit tests

St.❔	Category	Percentage	Covered / Total
🟢	Statements	82.64%	24205/29289
🟡	Branches	67.83%	10188/15020
🟡	Functions	77.76%	6530/8398
🟢	Lines	83.11%	23659/28466

Test suite run success

6775 tests passing in 786 suites.

Report generated by 🧪jest coverage report action from 6e6588a