src/app/admin/devices/components/DeviceModal.tsx (1)

23-24: 🏗️ Heavy lift

Replace the 1000-item preload with real async lookup.

Line 24 still preloads a large SoC list in the modal path, which can degrade open-time performance as data grows. Implement a search-driven async fetch instead of bulk loading and remove the TODO once done.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/app/admin/devices/components/DeviceModal.tsx` around lines 23 - 24, The
modal currently preloads 1000 SoC options via the call to
api.socs.options.useQuery({ limit: 1000 }) (socsQuery) which must be replaced
with a search-driven async fetch: remove the hardcoded limit call, add a
debounced search state in DeviceModal and call api.socs.options.useQuery or an
equivalent endpoint with the search term (e.g., { q: searchTerm, limit: 50 })
when the user types, wire those results into the select component (or use an
AsyncSelect pattern) to load options on demand, ensure loading/error states are
handled, and remove the TODO comment once complete.

sentry.edge.config.ts (1)

8-20: ⚡ Quick win

Trim explanatory inline comments in this block.

Keep only comments that capture non-obvious external constraints or business invariants.

Proposed cleanup

-// Only enable Sentry when explicitly configured for this deployment.
 if (process.env.NEXT_PUBLIC_ENABLE_SENTRY === 'true') {
   Sentry.init({
     dsn: 'https://85ca585e45005d8786e361c3456518bf@o74828.ingest.us.sentry.io/4509717207318529',

-    // Define how likely traces are sampled. Adjust this value in production, or use tracesSampler for greater control.
     tracesSampleRate: 1,

-    // Enable logs to be sent to Sentry
     enableLogs: true,

-    // Setting this option to true will print useful information to the console while you're setting up Sentry.
     debug: false,
   })
 }

As per coding guidelines `**/*.{ts,tsx,js,jsx}`: Comments should be rare, factual, and explain only non-obvious external constraints or business invariants.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@sentry.edge.config.ts` around lines 8 - 20, This block contains excessive
explanatory inline comments around the Sentry initialization; trim comments to
only those that convey non-obvious external constraints or business invariants.
In the if (process.env.NEXT_PUBLIC_ENABLE_SENTRY === 'true') { ... } /
Sentry.init({...}) section, remove or shorten comments for obvious configuration
keys (dsn, tracesSampleRate, enableLogs, debug) and keep a single factual
comment only if there is an external requirement (e.g., a deployment-level
mandate about NEXT_PUBLIC_ENABLE_SENTRY or mandatory DSN handling). Ensure
remaining comments are concise and factual.

src/app/pc-listings/components/filters/AsyncGpuFilterSelect.test.tsx (1)

68-84: ⚡ Quick win

Reduce non-mock inline string literals in this test.

Move assertion/label strings to fixture-derived constants so literals are limited to mocking data.

Proposed refactor

+const FILTER_LABEL = 'GPUs'
+const SELECTED_GPU = { id: 'gpu-selected', modelName: 'Radeon RX 7800 XT', brandName: 'AMD' }
+const OPTION_GPU = { id: 'gpu-1', modelName: 'RTX 4070', brandName: 'NVIDIA' }

 it('maps GPU option and selected labels', () => {
-  render(<AsyncGpuFilterSelect label="GPUs" value={['gpu-selected']} onChange={vi.fn()} />)
+  render(<AsyncGpuFilterSelect label={FILTER_LABEL} value={[SELECTED_GPU.id]} onChange={vi.fn()} />)

-  expect(screen.getByText('AMD Radeon RX 7800 XT')).toBeInTheDocument()
-  fireEvent.click(screen.getByRole('button', { name: 'GPUs multi-select' }))
-  expect(screen.getByText('NVIDIA RTX 4070')).toBeInTheDocument()
+  expect(
+    screen.getByText(`${SELECTED_GPU.brandName} ${SELECTED_GPU.modelName}`),
+  ).toBeInTheDocument()
+  fireEvent.click(screen.getByRole('button', { name: `${FILTER_LABEL} multi-select` }))
+  expect(screen.getByText(`${OPTION_GPU.brandName} ${OPTION_GPU.modelName}`)).toBeInTheDocument()
 })

As per coding guidelines `**/*.{test,spec}.{ts,tsx}`: Tests may use string literals only when mocking requires it.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/app/pc-listings/components/filters/AsyncGpuFilterSelect.test.tsx` around
lines 68 - 84, The test in AsyncGpuFilterSelect uses inline string literals for
labels and selected values; replace them with constants sourced from your test
fixtures/mock data instead of hard-coded strings. Update the test to import the
fixture constants (e.g., GPU_LABEL, GPU_OPTION_LABEL_NVIDIA,
GPU_OPTION_LABEL_AMD, GPU_SELECTED_ID) from the module used by setupApiMocks or
the API mock file, then use those constants in render(<AsyncGpuFilterSelect
label={GPU_LABEL} value={[GPU_SELECTED_ID]} .../>), in the screen.getByText
assertions, and in fireEvent.click(screen.getByRole('button', { name:
`${GPU_LABEL} multi-select` })); ensure setupApiMocks still provides the mock
data keys that the constants reference.

src/app/pc-listings/components/filters/AsyncCpuFilterSelect.test.tsx (1)

80-85: ⚡ Quick win

Reduce non-mock string literals in the test body.

Lines 80-85 use hard-coded UI strings in test title/props/assertions. Please switch to shared constants/test fixtures so literals remain only where mocking requires them.

As per coding guidelines "Tests may use string literals only when mocking requires it".

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/app/pc-listings/components/filters/AsyncCpuFilterSelect.test.tsx` around
lines 80 - 85, The test uses hard-coded UI strings; replace those literals with
shared test constants/fixtures: import canonical labels and option text
constants (e.g., CPU_LABEL, CPU_MULTI_SELECT_NAME, CPU_OPTION_INTEL,
CPU_OPTION_AMD or equivalent) and use them in the it() title, the render label
prop for AsyncCpuFilterSelect, and the assertions that call screen.getByText /
getByRole; update references to AsyncCpuFilterSelect, the test case name,
render(...) and the two expect(...) calls to use those imported constants so
strings only remain in mocks/fixtures.

src/components/popups/StopKillingGamesPopup.tsx (1)

12-12: ⚡ Quick win

Remove or replace the non-actionable TODO comment.

Line 12 adds a vague TODO that does not document a concrete invariant/constraint. Please remove it or replace it with a specific tracked action.

As per coding guidelines "Comments should be rare, factual, and explain only non-obvious external constraints or business invariants".

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/components/popups/StopKillingGamesPopup.tsx` at line 12, Remove the vague
TODO on line 12 in the StopKillingGamesPopup component: either delete the
non-actionable comment or replace it with a specific, factual note (e.g., a
one-line explanation of the invariant or a reference to a tracking ticket/PR) so
the comment documents a concrete constraint or next step; update the
StopKillingGamesPopup.tsx file's comment near the component declaration
accordingly.

src/server/api/routers/socs.ts (1)

22-22: ⚡ Quick win

Remove the TODO note or convert it into a tracked issue.

Inline TODOs here add noise without documenting a business invariant or external constraint.

As per coding guidelines, “Comments should be rare, factual, and explain only non-obvious external constraints or business invariants.”

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/server/api/routers/socs.ts` at line 22, Remove the inline comment "//
TODO: use paginate helpers" from src/server/api/routers/socs.ts; either delete
the TODO entirely or replace it with a short factual comment linking to a
created issue/PR (e.g., "See ISSUE-123 for paginate helper work") so the intent
is tracked externally rather than left as inline noise. Locate the exact TODO
string in the file and apply the change, ensuring no other logic is altered.

src/lib/cors.test.ts (1)

5-123: ⚡ Quick win

Replace repeated raw literals with named test constants.

This test now relies on many direct string literals outside mocking paths; move env keys/URLs/app-env values into constants to align with test conventions.

As per coding guidelines, “Tests may use string literals only when mocking requires it.”

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/lib/cors.test.ts` around lines 5 - 123, The tests use many repeated
string literals; define named test constants (e.g., ALLOWED_ORIGIN_EMUREADY =
'https://emuready.com', LOCALHOST_URL = 'http://localhost:3000', CI_ENV =
'true', PROD_ENV = 'production', TEST_APP_ENV = 'test') at the top of the file
and replace raw usages in calls to loadCors(), expectations for
getOriginFromUrl, getAllowedOrigins, and isAllowedRequestOrigin (and any
references to allowedOrigins or source strings) with those constants so the env
keys/URLs/app-env values are centralized and follow the test convention.

src/components/ui/form/async-multi-select/AsyncMultiSelect.test.tsx (1)

41-353: 🏗️ Heavy lift

Extract repeated UI query strings into named constants/fixtures.

This suite hardcodes many query strings (labels/placeholders/button names). Centralizing them in shared constants keeps tests resilient and aligns with repository test rules.

As per coding guidelines, "**/*.{test,spec}.{ts,tsx}: Tests may use string literals only when mocking requires it".

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/components/ui/form/async-multi-select/AsyncMultiSelect.test.tsx` around
lines 41 - 353, The tests hardcode many UI query strings (e.g., 'Devices
multi-select', 'Search devices...', 'Clear search', 'Retroid Pocket 5', 'AYN
Odin 2', 'Steam Deck OLED', etc.) which should be centralized; update the
AsyncMultiSelect.test.tsx to import and use a shared test-strings fixture (e.g.,
names like SELECT_BUTTON_LABEL, SEARCH_PLACEHOLDER, CLEAR_SEARCH_LABEL,
OPTION_RETROID_POCKET, OPTION_AYN_ODIN, SELECTED_STEAM_DECK) instead of literal
strings, replace all occurrences in the suite (including usages in renderSelect
calls and screen.getByRole/getByText/getByPlaceholderText assertions) and
add/update the fixture file exporting these constants so tests remain resilient
and follow the test string guideline.

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

docs/ERROR_HANDLING.md (2)

21-21: ⚡ Quick win

Consider consolidating the mirrored error formatter implementations.

The documentation notes that the error formatter is "mirrored in src/server/api/mobileContext.ts". Duplicate error formatting logic creates a maintenance burden and risks divergence between web and mobile error handling.

♻️ Suggested approach

Extract the error formatter into a shared utility function that both trpc.ts and mobileContext.ts can import, ensuring consistent error serialization across both contexts.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/ERROR_HANDLING.md` at line 21, The repo has duplicated error formatter
logic between src/server/api/trpc.ts and src/server/api/mobileContext.ts;
extract the common formatting into a shared utility (e.g., export a function
like formatAppError or getSerializableAppError wrapper) in a new module (or
existing src/lib/app-error-cause.ts), update trpc.ts and mobileContext.ts to
import and call that single function, and ensure the exported function returns
the same shape (data.appError) so both contexts remain behaviorally identical.

---

30-30: 💤 Low value

Clarify the consequence of violating the plain JSON constraint.

The constraint that cause fields must be plain JSON values is stated but doesn't explain what happens if a nested Error instance is passed. Adding the consequence (e.g., "will throw during serialization" or "will be omitted from the client payload") helps developers understand the severity.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@docs/ERROR_HANDLING.md` at line 30, The docs state that `cause` fields must
be plain JSON values but don’t explain the consequence; update the `cause`
documentation to explicitly say that passing a nested Error instance will break
serialization (e.g., cause JSON.stringify to fail or result in non-serializable
fields being stripped/omitted from the client payload) and add a short
remediation: convert Error instances to plain objects (message, name, stack)
before assigning to `cause`. Include the term `cause` and mention JSON
serialization/JSON.stringify so readers can locate the rule and the fix.

src/server/api/routers/games.test.ts (1)

79-98: ⚡ Quick win

Avoid as never in the caller helper.

This suppresses router-context drift in a brand-new test and can let broken mocks compile unnoticed. Please type the mock context to the caller contract instead of casting it away. As per coding guidelines **/*.{ts,tsx}: Do not use casts to hide type problems. Fix the underlying type issue.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/server/api/routers/games.test.ts` around lines 79 - 98, The helper
createCaller uses an unsafe cast ("prisma: prisma as never") which hides real
type mismatches; instead build and pass a properly typed context to
gamesRouter.createCaller by using the router's context parameter type (e.g. type
Ctx = Parameters<typeof gamesRouter.createCaller>[0]) and declare the context
object with session and prisma typed as Ctx['prisma'] (or import the actual
Context type used by the router) before calling
gamesRouter.createCaller(context); remove the as never cast and ensure
MockPrisma implements the expected prisma shape so TypeScript will validate the
mock.

src/features/human-verification/client/components/HumanVerificationProvider.test.tsx (1)

98-207: ⚡ Quick win

Add coverage for the new script-load failure branches.

These tests cover the ready/render paths well, but the provider now also rejects on VERIFICATION_SCRIPT_LOAD_TIMEOUT_MS and <Script onError>. A fake-timer timeout case plus a direct scriptProps[0].onError?.() case would protect the new user-visible failure modes.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In
`@src/features/human-verification/client/components/HumanVerificationProvider.test.tsx`
around lines 98 - 207, Add two tests to cover script-load failure branches: one
that uses fake timers to advance time past VERIFICATION_SCRIPT_LOAD_TIMEOUT_MS
and asserts the captured promise rejects with a script-load timeout error, and
another that directly invokes scriptProps[0].onError?.() after rendering with
window.turnstile unset and asserts the captured promise rejects with the script
load error; place these alongside the existing tests in
HumanVerificationProvider.test.tsx and reference
VERIFICATION_SCRIPT_LOAD_TIMEOUT_MS, scriptProps[0].onError, and the same
CaptureButton request-handling pattern to collect the promise for assertions.

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes