Skip to main content
Back to Categories
Category Suite: Autonomous

Full-Stack Evaluation Benchmark

API routes, database ORM queries, auth middleware.

Category Prompts & Benchmark Tasks (10)Zero-Shot & Multi-Turn Evals
#1 Next.js 16 API Route & Prisma ORM Handler
Difficulty: Hard

Write a type-safe Next.js API route handler with Prisma transactions, error handling, and rate limiting headers.

#2 JWT Authentication & RBAC Middleware
Difficulty: Hard

Implement a full-stack JWT authentication system featuring access/refresh token rotation, HTTP-only cookies, password hashing with bcrypt, and role-based access control middleware.

#3 Stripe Subscription Webhook Listener
Difficulty: Hard

Build an enterprise Stripe webhook integration handler supporting signature verification, customer subscription state synchronization, and database transaction updates.

#4 Real-time WebSockets Chat Backend Server
Difficulty: Hard

Create a Node.js WebSocket server supporting multi-room chat channels, client heartbeat ping/pong, message persistence, and presence tracking.

#5 File Upload Manager with S3 Presigned URLs
Difficulty: Medium

Build a secure file upload pipeline with client-side image compression, S3 presigned URL generation, file type validation, and database asset metadata creation.

#6 REST API Rate Limiter & Token Bucket
Difficulty: Hard

Implement a Redis-backed token bucket rate limiting middleware with sliding window algorithms, custom quota tier overrides, and standard HTTP rate limit headers.

#7 GraphQL Schema & Resolver Suite
Difficulty: Hard

Write a complete GraphQL server schema with type definitions, queries, mutations, dataloader batching to solve N+1 problems, and field-level authorization policies.

#8 Multi-Tenant Database Row Level Security (RLS)
Difficulty: Hard

Implement a multi-tenant database isolation architecture using PostgreSQL Row Level Security policies, tenant context middleware, and automated migration scripts.

#9 Background Job Queue & Task Worker
Difficulty: Hard

Build an asynchronous background task processing system with BullMQ/Redis queues, failure retry strategies, task progress reporting, and dead-letter queues.

#10 Server-Sent Events (SSE) Streaming API
Difficulty: Medium

Create a Next.js Server-Sent Events route handler that streams real-time AI token generation chunks to the client with reconnection event id tracking.