gh-greyhaven-ai-claude-code-config-grey-haven-plugins-research/skills/tanstack-patterns/SKILL.md

name, description

name	description
grey-haven-tanstack-patterns	Apply Grey Haven's TanStack ecosystem patterns - Router file-based routing, Query data fetching with staleTime, and Start server functions. Use when building React applications with TanStack Start.

Grey Haven TanStack Patterns

Follow Grey Haven Studio's patterns for TanStack Start, Router, and Query in React 19 applications.

TanStack Stack Overview

Grey Haven uses the complete TanStack ecosystem:

• TanStack Start: Full-stack React framework with server functions
• TanStack Router: Type-safe file-based routing with loaders
• TanStack Query: Server state management with caching
• TanStack Table (optional): Data grids and tables
• TanStack Form (optional): Type-safe form handling

Critical Patterns

1. File-Based Routing Structure

src/routes/
├── __root.tsx              # Root layout (wraps all routes)
├── index.tsx               # Homepage (/)
├── _authenticated/         # Protected routes group (underscore prefix)
│   ├── _layout.tsx        # Auth layout wrapper
│   ├── dashboard.tsx      # /dashboard
│   └── settings/
│       └── index.tsx      # /settings
└── users/
    ├── index.tsx          # /users
    └── $userId.tsx        # /users/:userId (dynamic param)

Key conventions:

• __root.tsx - Root layout with QueryClient provider
• _authenticated/ - Protected route groups (underscore prefix)
• _layout.tsx - Layout wrapper for route groups
• $param.tsx - Dynamic route parameters

2. TanStack Query Defaults

const queryClient = new QueryClient({
  defaultOptions: {
    queries: {
      staleTime: 60000, // 1 minute default
      retry: 1,
      refetchOnWindowFocus: false,
    },
  },
});

3. Query Key Patterns

// ✅ CORRECT - Specific to general
queryKey: ["user", userId]
queryKey: ["users", { tenantId, page: 1 }]
queryKey: ["organizations", orgId, "teams"]

// ❌ WRONG
queryKey: [userId]                    // Missing resource type
queryKey: ["getUser", userId]        // Don't include function name
queryKey: [{ id: userId }]           // Object first is confusing

4. Server Functions with Multi-Tenant

// ALWAYS include tenant_id parameter
export const getUserById = createServerFn("GET", async (
  userId: string,
  tenantId: string
) => {
  const user = await db.query.users.findFirst({
    where: and(
      eq(users.id, userId),
      eq(users.tenant_id, tenantId) // Multi-tenant isolation!
    ),
  });

  if (!user) throw new Error("User not found");
  return user;
});

5. Route Loaders

export const Route = createFileRoute("/_authenticated/dashboard")({
  // Loader fetches data on server before rendering
  loader: async ({ context }) => {
    const tenantId = context.session.tenantId;
    return await getDashboardData(tenantId);
  },
  component: DashboardPage,
});

function DashboardPage() {
  const data = Route.useLoaderData(); // Type-safe loader data
  return <div>...</div>;
}

6. Mutations with Cache Invalidation

const mutation = useMutation({
  mutationFn: (data: UserUpdate) => updateUser(userId, data),
  // Always invalidate queries after mutation
  onSuccess: () => {
    queryClient.invalidateQueries({ queryKey: ["user", userId] });
  },
});

Caching Strategy

Grey Haven uses these staleTime defaults:

Data Type	staleTime	Use Case
Auth data	5 minutes	User sessions, tokens
User profiles	1 minute	User details
Lists	1 minute	Data tables, lists
Static/config	10 minutes	Settings, configs
Realtime	0 (always refetch)	Notifications

const STALE_TIMES = {
  auth: 5 * 60 * 1000,        // 5 minutes
  user: 1 * 60 * 1000,        // 1 minute
  list: 1 * 60 * 1000,        // 1 minute
  static: 10 * 60 * 1000,     // 10 minutes
  realtime: 0,                // Always refetch
};

Supporting Documentation

All supporting files are under 500 lines per Anthropic best practices:

• examples/ - Complete code examples

  • router-patterns.md - File-based routing, layouts, navigation
  • query-patterns.md - Queries, mutations, infinite queries
  • server-functions.md - Creating and using server functions
  • advanced-patterns.md - Dependent queries, parallel queries, custom hooks
  • INDEX.md - Examples navigation

• reference/ - Configuration references

  • router-config.md - Router setup and configuration
  • query-config.md - QueryClient configuration
  • caching-strategy.md - Detailed caching patterns
  • multi-tenant.md - Multi-tenant patterns with RLS
  • INDEX.md - Reference navigation

• templates/ - Copy-paste ready templates

  • root-route.tsx - Root layout template
  • auth-layout.tsx - Protected layout template
  • page-route.tsx - Basic page route template
  • server-function.ts - Server function template
  • custom-hook.ts - Custom query hook template

• checklists/ - Pre-PR validation

  • tanstack-checklist.md - TanStack patterns checklist

When to Apply This Skill

Use this skill when:

• Building TanStack Start applications
• Implementing routing with TanStack Router
• Managing server state with TanStack Query
• Creating server functions for data fetching
• Optimizing query performance with caching
• Implementing multi-tenant data access
• Setting up authentication flows with route protection
• Building data-heavy React applications

Template Reference

These patterns are from Grey Haven's production template:

• cvi-template: TanStack Start + Router + Query + React 19

Critical Reminders

1. staleTime: Default 60000ms (1 minute) for queries
2. Query keys: Specific to general (["user", userId], not [userId])
3. Server functions: Always include tenant_id parameter
4. Multi-tenant: Filter by tenant_id in all server functions
5. Loaders: Use for server-side data fetching before render
6. Mutations: Invalidate queries after successful mutation
7. Prefetching: Use for performance on hover/navigation
8. Error handling: Always handle error state in queries
9. RLS: Server functions use RLS-enabled database connection
10. File-based routing: Underscore prefix (_) for route groups/layouts