Skip to main content

EGI SKU Standard

Standardized SKU (Stock Keeping Unit) format for versioning systems and releases across all platforms and system types.

Overview

The EGI SKU standard provides a consistent way to identify, version, and track released systems across different platforms and system types. Every released component should have a unique SKU that captures its version, purpose, platform, and build date.

Format

X.Y.Z.TYPE.PLATFORM.YYYYMMDD

Components

  • X.Y.Z - Semantic version (major.minor.patch)
  • TYPE - System type code (API, BOT, RAG, etc.)
  • PLATFORM - Deployment platform code (EKS, GKE, LM, etc.)
  • YYYYMMDD - Build/release date

Quick Examples

1.0.0.API.EKS.20250127    # API on AWS EKS, version 1.0.0
2.5.1.BOT.GKE.20250115 # Bot worker on Google GKE, version 2.5.1
1.2.0.RAG.LM.20250201 # RAG service on AWS Lambda, version 1.2.0
3.0.0.WA.VE.20250110 # Web app on Vercel, version 3.0.0

Philosophy

The SKU standard is designed to:

  1. Provide traceability - Know exactly what version is running where
  2. Enable automation - Machines can parse and validate SKUs
  3. Support multi-platform - Same code, different platforms, different SKUs
  4. Track deployments - Build date helps correlate with CI/CD runs
  5. Enforce semantic versioning - Clear version progression

Why Use SKUs?

  • Deployment tracking - Quickly identify what's running in each environment
  • Rollback safety - Know exactly which version to revert to
  • Change management - Tie deployments to specific builds and dates
  • Audit compliance - Track what was deployed when and where
  • CI/CD integration - Automate version management and validation

Documentation Structure

This section covers:

Getting Started

  1. Review the naming conventions to understand the format
  2. Find your platform code and system type
  3. Check out examples for similar deployments
  4. Use the tooling to generate and validate SKUs
  5. Follow release conventions as you evolve your service

Quick Reference

Common Platform Codes

CodePlatformUse For
EKSAWS Elastic Kubernetes ServiceProduction containerized workloads on AWS
GKEGoogle Kubernetes EngineProduction containerized workloads on GCP
LMAWS LambdaServerless functions
VEVercelNext.js and frontend deployments
CFCloudflare Workers/PagesEdge compute and static sites

Common System Types

CodeDescriptionUse For
APIREST/GraphQL API serviceBackend APIs and microservices
BOTBackground worker/botQueue workers, cron jobs, automation
RAGRAG systemRetrieval-augmented generation systems
WAWeb applicationFrontend web apps (SPAs, SSR)
SDKClient SDKPython/JavaScript packages for developers

See the full lists in Platform Identifiers and System Types.

Support

For questions or to propose new platform codes or system types:

  • submit a PR to this repo
  • contact the platform team
  • see Update Process for contribution guidelines