Review Tanstack Router

Review TanStack Router

Review the TanStack Router setup in front-app for alignment with current official best practices - route-tree generation, type-safe search params, loading/preloading strategy, and developer experience on a React 19 SPA deployed via Cloudflare. Your reply must be a plan of suggested changes: concise, actionable, structured - not only prose.

Invocation

Text after the slash command is additional scope/focus (e.g. "search params", "code splitting") - narrow the review accordingly.

Ground truth (mandatory)

Your pre-trained knowledge of TanStack Router may be outdated (frequent minor releases). Do not draft suggestions from memory alone.

1. Resolve "TanStack Router" via the Context7 MCP (resolve-library-id → query-docs): file-based routing conventions, createFileRoute/route APIs at the installed version, search-param validation with Zod, code-based splitting/lazyRouteComponent patterns, devtools.
2. Cross-check with local skill .agents/skills/tanstack-router/SKILL.md for repo-relevant depth.
3. For anything Context7 lacks, use Firecrawl search/scrape restricted to the official domain (tanstack.com/router) - guides and changelog.
4. Version currency: catalog entries (@tanstack/react-router, @tanstack/router-plugin, @tanstack/router-cli, -devtools) in pnpm-workspace.yaml vs latest stable; plugin and runtime version skew is a classic drift point.
5. Cite the retrieved source next to every finding; label anything unverifiable as Unverified.

Scope artifacts

• Route files under apps/front-app/src/routes/ and generated route tree artifact
• apps/front-app/tsr.config.json, router-plugin options in apps/front-app/vite.config.ts
• CLI usage (@tanstack/router-cli generate task) in package scripts; generated-file ignore policy (oxlint/knip/git)
• Devtools wiring (@tanstack/react-router-devtools)

Analysis axes

• Route-tree generation: generator config current; generated file committed/ignored per repo policy and excluded from lint/format noise; regeneration wired into dev/build reliably.
• Type safety: validateSearch with Zod schemas shared with DTOs where applicable; fully typed Link/navigate usage; no as any escapes around params.
• Loading & data: loaders used appropriately vs TanStack Query composition (loader awaiting queryClient.ensureQueryData or deferring); Suspense/error boundaries per route; no waterfall-prone sequential awaits.
• Code splitting & preloading: route-level lazy chunks; preload intent configuration sensible; default gcTime/stale-time interplay with query cache understood.
• Devtools: gated out of production builds; version aligned with runtime.
• Version currency: breaking changes between catalog pin and latest noted when relevant to used APIs.

DX & AI-agentic workflow

Verify agent-friendliness: adding a route is a documented repeatable procedure (nested AGENTS.md); generated artifacts never hand-edited; type errors from the route tree actionable via turbo run check-types --filter=front-app.

Steps

1. Collect ground truth before reading code.
2. Read tsr config + vite plugin options + all route files.
3. Walk each analysis axis; note findings or explicit one-line "no issues".
4. Compose the plan grouped Critical / Improvements / Optional with what, where, why, and source citations.

Output format

1. Critical - broken generation, untyped/unvalidated params reaching logic, stale plugin/runtime mismatch.
2. Improvements - loader/splitting/search-param alignment with current guidance.
3. Optional - nice-to-haves; prefix pure polish with Nit:.

Read-only review: produce the plan only; implement nothing unless explicitly asked afterwards.