A collection of skills for AI coding agents. Skills are packaged instructions and scripts that extend agent capabilities.
Skills follow the Agent Skills format.
NestJS best practices and patterns for building scalable, maintainable backend applications. Contains 26 rules across 13 categories covering security, architecture, performance, validation, database operations, authentication, and advanced patterns.
Use when:
- Writing new NestJS modules, controllers, or services
- Implementing authentication and authorization
- Setting up Prisma database operations
- Creating DTOs and validation pipes
- Adding middleware or guards
- Implementing error handling and logging
- Reviewing NestJS code for best practices
- Refactoring existing NestJS applications
Categories covered:
- Security (CRITICAL) - CORS, dependency audits, security headers
- Performance (HIGH) - Redis caching, compression
- Architecture (HIGH) - Feature modules, naming conventions, event-driven architecture
- Error Handling (HIGH) - Exception filters, structured logging
- Validation (CRITICAL) - DTOs, ValidationPipe, custom pipes
- Database (CRITICAL) - Parameterized queries, Prisma v7 patterns
- Authentication (CRITICAL) - Guards, password hashing (argon2/bcrypt)
- API (MEDIUM) - Swagger, cursor pagination
- Configuration (CRITICAL) - Environment variables, secrets management
- Testing (MEDIUM) - Unit tests, test coverage
- Deployment (MEDIUM) - Health checks, graceful shutdown
- Middleware (MEDIUM) - Rate limiting, compression
- Advanced (HIGH) - Scheduled tasks, lazy loading
Technology stack:
- NestJS with Express/Fastify adapter
- Bun runtime (native crypto, scheduling)
- Prisma v7 (sql template tag, interactive transactions)
- EventEmitter2 for event-driven architecture
npx add-skill xirothedev/agent-skillsSkills are automatically available once installed. The agent will use them when relevant tasks are detected.
Examples:
Review my NestJS service for security issues
Create authentication guards for my NestJS app
Help me implement cursor pagination
Each skill contains:
SKILL.md- Instructions for the agentrules/- Individual rule files with agent-friendly formatAGENTS.md- Compiled documentation (generated)README.md- Skill-specific documentationLICENSE- License information
MIT