convex-best-practices

Convex Best Practices

Build production-ready Convex applications by following established patterns for function organization, query optimization, validation, TypeScript usage, and error handling.

Documentation Sources

Before implementing, do not assume; fetch the latest documentation:

• Primary: https://docs.convex.dev/understanding/best-practices/
• Error Handling: https://docs.convex.dev/functions/error-handling
• Write Conflicts: https://docs.convex.dev/error#1
• For broader context: https://docs.convex.dev/llms.txt

Instructions

The Zen of Convex

1. Convex manages the hard parts - Let Convex handle caching, real-time sync, and consistency
2. Functions are the API - Design your functions as your application's interface
3. Schema is truth - Define your data model explicitly in schema.ts
4. TypeScript everywhere - Leverage end-to-end type safety
5. Queries are reactive - Think in terms of subscriptions, not requests

Function Organization

Organize your Convex functions by domain:

// convex/users.ts - User-related functions
import { query, mutation } from './_generated/server'
import { v } from 'convex/values'

export const get = query({
  args: { userId: v.id('users') },
  returns: v.union(
    v.object({
      _id: v.id('users'),
      _creationTime: v.number(),
      name: v.string(),
      email: v.string(),
    }),
    v.null(),
  ),
  handler: async (ctx, args) => {
    return await ctx.db.get(args.userId)
  },
})

Argument and Return Validation

Always define validators for arguments AND return types:

export const createTask = mutation({
  args: {
    title: v.string(),
    description: v.optional(v.string()),
    priority: v.union(v.literal('low'), v.literal('medium'), v.literal('high')),
  },
  returns: v.id('tasks'),
  handler: async (ctx, args) => {
    return await ctx.db.insert('tasks', {
      title: args.title,
      description: args.description,
      priority: args.priority,
      completed: false,
      createdAt: Date.now(),
    })
  },
})

Query Patterns

Use indexes instead of filters for efficient queries:

// Schema with index
export default defineSchema({
  tasks: defineTable({
    userId: v.id('users'),
    status: v.string(),
    createdAt: v.number(),
  })
    .index('by_user', ['userId'])
    .index('by_user_and_status', ['userId', 'status']),
})

// Query using index
export const getTasksByUser = query({
  args: { userId: v.id('users') },
  returns: v.array(
    v.object({
      _id: v.id('tasks'),
      _creationTime: v.number(),
      userId: v.id('users'),
      status: v.string(),
      createdAt: v.number(),
    }),
  ),
  handler: async (ctx, args) => {
    return await ctx.db
      .query('tasks')
      .withIndex('by_user', (q) => q.eq('userId', args.userId))
      .order('desc')
      .collect()
  },
})

Error Handling

Use ConvexError for user-facing errors:

import { ConvexError } from 'convex/values'

export const updateTask = mutation({
  args: {
    taskId: v.id('tasks'),
    title: v.string(),
  },
  returns: v.null(),
  handler: async (ctx, args) => {
    const task = await ctx.db.get(args.taskId)

    if (!task) {
      throw new ConvexError({
        code: 'NOT_FOUND',
        message: 'Task not found',
      })
    }

    await ctx.db.patch(args.taskId, { title: args.title })
    return null
  },
})

Avoiding Write Conflicts (Optimistic Concurrency Control)

Convex uses OCC. Follow these patterns to minimize conflicts:

// GOOD: Make mutations idempotent
export const completeTask = mutation({
  args: { taskId: v.id('tasks') },
  returns: v.null(),
  handler: async (ctx, args) => {
    const task = await ctx.db.get(args.taskId)

    // Early return if already complete (idempotent)
    if (!task || task.status === 'completed') {
      return null
    }

    await ctx.db.patch(args.taskId, {
      status: 'completed',
      completedAt: Date.now(),
    })
    return null
  },
})

// GOOD: Patch directly without reading first when possible
export const updateNote = mutation({
  args: { id: v.id('notes'), content: v.string() },
  returns: v.null(),
  handler: async (ctx, args) => {
    // Patch directly - ctx.db.patch throws if document doesn't exist
    await ctx.db.patch(args.id, { content: args.content })
    return null
  },
})

// GOOD: Use Promise.all for parallel independent updates
export const reorderItems = mutation({
  args: { itemIds: v.array(v.id('items')) },
  returns: v.null(),
  handler: async (ctx, args) => {
    const updates = args.itemIds.map((id, index) =>
      ctx.db.patch(id, { order: index }),
    )
    await Promise.all(updates)
    return null
  },
})

TypeScript Best Practices

import { Id, Doc } from './_generated/dataModel'

// Use Id type for document references
type UserId = Id<'users'>

// Use Doc type for full documents
type User = Doc<'users'>

// Define Record types properly
const userScores: Record<Id<'users'>, number> = {}

Internal vs Public Functions

// Public function - exposed to clients
export const getUser = query({
  args: { userId: v.id('users') },
  returns: v.union(
    v.null(),
    v.object({
      /* ... */
    }),
  ),
  handler: async (ctx, args) => {
    // ...
  },
})

// Internal function - only callable from other Convex functions
export const _updateUserStats = internalMutation({
  args: { userId: v.id('users') },
  returns: v.null(),
  handler: async (ctx, args) => {
    // ...
  },
})

Examples

Complete CRUD Pattern

// convex/tasks.ts
import { query, mutation } from './_generated/server'
import { v } from 'convex/values'
import { ConvexError } from 'convex/values'

const taskValidator = v.object({
  _id: v.id('tasks'),
  _creationTime: v.number(),
  title: v.string(),
  completed: v.boolean(),
  userId: v.id('users'),
})

export const list = query({
  args: { userId: v.id('users') },
  returns: v.array(taskValidator),
  handler: async (ctx, args) => {
    return await ctx.db
      .query('tasks')
      .withIndex('by_user', (q) => q.eq('userId', args.userId))
      .collect()
  },
})

export const create = mutation({
  args: {
    title: v.string(),
    userId: v.id('users'),
  },
  returns: v.id('tasks'),
  handler: async (ctx, args) => {
    return await ctx.db.insert('tasks', {
      title: args.title,
      completed: false,
      userId: args.userId,
    })
  },
})

export const update = mutation({
  args: {
    taskId: v.id('tasks'),
    title: v.optional(v.string()),
    completed: v.optional(v.boolean()),
  },
  returns: v.null(),
  handler: async (ctx, args) => {
    const { taskId, ...updates } = args

    // Remove undefined values
    const cleanUpdates = Object.fromEntries(
      Object.entries(updates).filter(([_, v]) => v !== undefined),
    )

    if (Object.keys(cleanUpdates).length > 0) {
      await ctx.db.patch(taskId, cleanUpdates)
    }
    return null
  },
})

export const remove = mutation({
  args: { taskId: v.id('tasks') },
  returns: v.null(),
  handler: async (ctx, args) => {
    await ctx.db.delete(args.taskId)
    return null
  },
})

Best Practices

• Never run npx convex deploy unless explicitly instructed
• Never run any git commands unless explicitly instructed
• Always define return validators for functions
• Use indexes for all queries that filter data
• Make mutations idempotent to handle retries gracefully
• Use ConvexError for user-facing error messages
• Organize functions by domain (users.ts, tasks.ts, etc.)
• Use internal functions for sensitive operations
• Leverage TypeScript's Id and Doc types

Common Pitfalls

1. Using filter instead of withIndex - Always define indexes and use withIndex
2. Missing return validators - Always specify the returns field
3. Non-idempotent mutations - Check current state before updating
4. Reading before patching unnecessarily - Patch directly when possible
5. Not handling null returns - Document IDs might not exist

References

• Convex Documentation: https://docs.convex.dev/
• Convex LLMs.txt: https://docs.convex.dev/llms.txt
• Best Practices: https://docs.convex.dev/understanding/best-practices/
• Error Handling: https://docs.convex.dev/functions/error-handling
• Write Conflicts: https://docs.convex.dev/error#1