Software Backend

What it does

This skill enables AI agents to design, implement, and review production-grade backend services. It automates server-side logic, API development (REST, GraphQL, tRPC), database interactions (Postgres with pooling), authentication, caching, observability using OpenTelemetry, error handling, testing, and deployment. The skill emphasizes type-safe boundaries, idempotency, RFC 9457 errors, structured logs, timeouts, and rate limiting for scalable backend systems.

When to use it

• Developing REST, GraphQL, or tRPC APIs.
• Modeling data schemas and managing database migrations.
• Implementing authentication and authorization using OIDC/OAuth, sessions, or JWTs.
• Adding features like validation, error handling, rate limiting, caching, and background jobs to backend services.
• Preparing a backend service for production deployment, including observability and runbooks.

Key capabilities

• API development with Fastify, Express, NestJS, Hono, or Apollo Server.
• Database interaction using PostgreSQL with pooling and migrations.
• Authentication implementation utilizing OIDC/OAuth, sessions, and JWTs.
• Validation using Zod, Pydantic, or similar libraries at the API boundary.
• Caching with Redis (or managed services).
• Background job processing with BullMQ or platform queues.
• Observability through structured logs and OpenTelemetry integration.
• Scaffolding complete working implementations for domain logic like fraud rules, audit logging, and webhook handlers.

Example prompts

• "Create a REST API endpoint to handle user registration using Fastify."
• "Generate a GraphQL schema for managing product inventory with Prisma as the ORM."
• "Implement authentication with JWTs and httpOnly cookies for a tRPC API."

Tips & gotchas

• The skill defaults towards specific technologies (e.g., PostgreSQL, Redis, Fastify). Consider these choices based on your team's expertise or project constraints.
• Validation should be implemented at the API boundary to ensure type safety and prevent errors deep within the application logic.
• This skill is not intended for frontend-only concerns, infrastructure provisioning, or SQL query optimization; use specialized skills for those tasks instead.