Comprehensive guide to all 40 built-in service templates for microservices and custom backends

nself provides 40+ production-ready service templates spanning 10 programming languages. Each template includes Docker configuration, health checks, CORS support, graceful shutdown, and follows language-specific best practices.

# Enable custom services
SERVICES_ENABLED=true

# Add a service using a template
CS_1=api:fastapi:3001:/api

# Build and start
nself build && nself start

• JavaScript/TypeScript: 19 templates (Node, Express, Fastify, NestJS, Hono, Socket.IO, BullMQ, Temporal, Bun, Deno, tRPC)
• Python: 7 templates (Flask, FastAPI, Django REST, Celery, Ray, AI Agents)
• Go: 4 templates (Gin, Echo, Fiber, gRPC)
• Other Languages: 10 templates (Rust, Java, C#, C++, Ruby, Elixir, PHP, Kotlin, Swift)

CS_N=name:template:port:route

• name: Service identifier (alphanumeric + hyphens)
• template: Template from the list below
• port: Container port (optional, auto-assigns)
• route: URL path prefix or domain (optional)

• node-js - Plain Node.js HTTP server (JavaScript)
• node-ts - Plain Node.js HTTP server (TypeScript)

• express-js - Express.js web framework (JavaScript)
• express-ts - Express.js web framework (TypeScript)
• fastify-js - High-performance Fastify framework (JavaScript)
• fastify-ts - High-performance Fastify framework (TypeScript)
• hono-js - Ultra-light edge-optimized framework (JavaScript)
• hono-ts - Ultra-light edge-optimized framework (TypeScript)

• nest-js - NestJS enterprise framework (JavaScript)
• nest-ts - NestJS enterprise framework (TypeScript)

• socketio-js - Socket.IO real-time bidirectional communication (JavaScript)
• socketio-ts - Socket.IO real-time bidirectional communication (TypeScript)

• bullmq-js - Redis-backed job queues with BullMQ (JavaScript)
• bullmq-ts - Redis-backed job queues with BullMQ (TypeScript)

• temporal-js - Temporal workflow orchestration (JavaScript)
• temporal-ts - Temporal workflow orchestration (TypeScript)

• bun - Bun runtime server (JavaScript)
• deno - Deno runtime server (TypeScript)

• trpc - Type-safe RPC framework (TypeScript only)

# Express API with TypeScript
CS_1=api:express-ts:3001:/api

# Real-time WebSocket service
CS_2=chat:socketio-ts:3002

# Background job processor
CS_3=jobs:bullmq-ts:3003

# High-performance API
CS_4=fast:fastify-ts:3004:/v1

# Enterprise microservice
CS_5=service:nest-ts:3005

• flask - Lightweight Flask microframework
• fastapi - Async type-hinted FastAPI framework
• django-rest - Django REST Framework APIs

• celery - Distributed task queue with Redis backend
• ray - Distributed ML compute and model serving

• agent-llm - LLM agent orchestration starter with OpenAI integration
• agent-data - Data-centric agent with pandas, scikit-learn, and DuckDB

# FastAPI microservice
CS_1=api:fastapi:8001:/api

# Background task processor
CS_2=tasks:celery:8002

# ML model serving
CS_3=ml:ray:8003:/models

# AI agent service
CS_4=agent:agent-llm:8004:/chat

# Data processing service
CS_5=data:agent-data:8005:/process

• gin - High-performance Gin web framework
• echo - Minimal Echo API framework
• fiber - Express-inspired, speed-focused Fiber framework
• grpc - Official Go gRPC implementation with Protocol Buffers

# High-performance API
CS_1=api:gin:8001:/api

# Minimal microservice
CS_2=service:echo:8002

# Fast web service
CS_3=web:fiber:8003

# gRPC service
CS_4=grpc:grpc:50051

• rust/axum - Modern async Rust web framework with Tokio

• java/spring-boot - Enterprise Java Spring Boot framework

• csharp/aspnet - ASP.NET Core Web API

• cpp/oatpp - Modern C++ web framework with high performance

• ruby/rails - Ruby on Rails in API mode

• elixir/phoenix - Productive Elixir Phoenix framework

• php/laravel - Laravel PHP framework

• kotlin/ktor - Asynchronous Kotlin Ktor framework

• swift/vapor - Swift Vapor server-side framework

# Rust high-performance API
CS_1=api:rust/axum:8001

# Java enterprise service
CS_2=enterprise:java/spring-boot:8002

# C# .NET API
CS_3=dotnet:csharp/aspnet:8003

# Ruby on Rails API
CS_4=rails:ruby/rails:8004

# Elixir Phoenix service
CS_5=phoenix:elixir/phoenix:8005

Every template includes:

• Multi-stage builds for smaller images
• Non-root container users
• Optimized layer caching
• Health check endpoints

• Security headers (X-Frame-Options, X-Content-Type-Options, etc.)
• CORS configuration
• Input validation patterns
• No secrets in container images

• Health check endpoints at /health
• Structured logging support
• Environment-based configuration
• Version information in responses

• Graceful shutdown handling
• Connection pooling where applicable
• Production-optimized dependencies
• Efficient resource utilization

• Template variable placeholders ({{SERVICE_NAME}}, {{SERVICE_PORT}}, etc.)
• Consistent file structure across languages
• Clear documentation comments
• Ready-to-customize code

Create your own templates in src/templates/services/custom/:

# Use custom template
CS_1=myservice:custom:8001

# nself will look for:
# src/templates/services/custom/Dockerfile.template
# src/templates/services/custom/[source files]

# Per-service environment variables
CS_1_ENV=NODE_ENV=production,LOG_LEVEL=info,API_KEY=${API_KEY}

# Health check customization
CS_1_HEALTH_PATH=/api/health
CS_1_HEALTH_TIMEOUT=30s

# Internal service (no external access)
CS_1=internal:fastapi:8001

# Public service with domain
CS_2=api:fastapi:8002:api.example.com

# Service with custom route
CS_3=v1:fastapi:8003:/v1/api

# Multiple instances of the same service
CS_1=api-1:fastapi:8001:/api
CS_2=api-2:fastapi:8002:/api
CS_3=api-3:fastapi:8003:/api

• All 40 templates are now fully validated and production-ready
• Template names are standardized (use hyphens, not underscores)
• TypeScript versions available for most JavaScript frameworks

• nodejs-raw → node-js
• nodejs-raw-ts → node-ts
• Removed incomplete templates from previous versions

• High Performance: fastify-ts, gin, rust/axum
• Enterprise: nest-ts, java/spring-boot
• Rapid Development: fastapi, express-ts
• Real-time: socketio-ts, elixir/phoenix
• Background Jobs: bullmq-ts, celery

# Use environment-specific settings
CS_1_ENV=NODE_ENV=${ENV},LOG_LEVEL=debug,DATABASE_URL=${DATABASE_URL}

# Configure appropriate timeouts
CS_1_HEALTH_TIMEOUT=30s
CS_1_HEALTH_INTERVAL=30s

# Set appropriate memory limits
CS_1_MEMORY=512m
CS_1_CPUS=0.5

# List available templates
find src/templates/services/ -name "Dockerfile.template"

# Verify template structure
ls src/templates/services/js/express-ts/

# Check template validation
nself validate

# Review build logs
nself logs CS_1

# Check service logs
nself logs [service-name]

# Verify port availability
nself status

• Configuration Reference - Detailed service configuration
• Examples - Real-world usage examples
• Architecture - How services fit into nself
• Template Validation - Quality assurance report

Need a new template? Submit a feature request or contribute your own template following the existing patterns.