SDK Release & Git Workflows

Comprehensive guides for releasing the Kailash Python SDK to PyPI, managing CI/CD pipelines, deploying documentation, and Git workflow best practices.

Overview

SDK release and development infrastructure patterns for:

• PyPI package publishing (TestPyPI + production)
• CI/CD pipelines (GitHub Actions)
• Multi-platform wheel building
• Documentation deployment
• Git workflows and branching strategies
• Multi-package version coordination

Reference Documentation

Deployment Lifecycle

• deployment-onboarding - SDK release onboarding process
  • Codebase analysis (packages, build system, CI, docs)
  • Structured questions for human architect
  • Research current PyPI/CI best practices
  • Create deployment-config.md
• release-runbook - SDK release runbook (detailed procedures)
  • Version locations for all packages (pyproject.toml + init.py)
  • SDK dependency pin rules
  • Version consistency verification commands
  • Pre-release, build, publish, post-release step-by-step procedures
• deployment-packages - Package release workflow
  • PyPI and GitHub release process
  • Multi-package coordination and publish order
  • Version bumping and changelog
  • CI-triggered releases
  • TestPyPI validation
  • Rollback procedures
• deployment-ci - CI/CD infrastructure
  • GitHub Actions workflows for Python packages
  • Multi-platform wheel building
  • Test matrix (Python versions x OS)
  • Tag-triggered publishing pipeline
  • Documentation deployment (ReadTheDocs, GitHub Pages)
  • Self-hosted runner management

Docker

• deployment-docker-quick - Docker deployment patterns
  • Dockerfile setup for Kailash apps
  • Docker Compose configurations
  • Multi-stage builds
  • Health checks

Kubernetes

• deployment-kubernetes-quick - Kubernetes deployment patterns
  • Deployment manifests
  • Service configuration
  • Scaling strategies

Git Workflow

• git-workflow-quick - Git workflow best practices
  • Branching strategies
  • Commit conventions
  • Pull request workflow
  • Code review process
  • Release management
  • Hotfix procedures

GitHub Management

• github-management-patterns - GitHub project and issue management
  • Issue templates (User Story, Bug, Technical Task)
  • Story points and estimation
  • Project board organization
  • Label system

Project Management

• project-management - Project management architecture
  • Dual-tracking system overview
  • GitHub Issues vs Local Todos
  • Agent coordination flow
  • Sprint management
• todo-github-sync - Todo ↔ GitHub issues sync patterns
  • Naming conventions (Story X format)
  • Workflow for creating, starting, completing stories
  • Sub-issue management
  • Label system
  • Periodic sync checklists
  • Agent coordination (todo-manager ↔ gh-manager)

SDK Release Patterns

Release Flow

Version bump + CHANGELOG → Build wheels → TestPyPI → Verify → PyPI → GitHub Release → Deploy Docs

CI-Triggered Release (Preferred)

git tag v1.2.3 → push tag → CI builds wheels → CI tests → CI publishes TestPyPI → CI publishes PyPI → CI creates GitHub Release

Multi-Package Release Order

kailash (core)
  ↓
kailash-dataflow (depends on core)
kailash-nexus (depends on core)
kailash-kaizen (depends on core)

Git Workflow Patterns

Branch Strategy

main (production)
  ↓
develop (integration)
  ↓
feature/* (new features)
hotfix/* (urgent fixes)
release/* (release prep)

Commit Conventions

feat: Add user authentication workflow
fix: Resolve async runtime threading issue
docs: Update DataFlow integration guide
test: Add cycle workflow test cases
chore: Bump version to 0.9.25

Critical Rules

SDK Release

• Run full test suite before any release
• TestPyPI validation required for major/minor releases
• Wheel-only publishing for proprietary code
• Version consistency across all sub-packages
• Publish in dependency order (core first)
• Security review before every publish
• NEVER commit PyPI tokens to source
• NEVER publish with failing CI

Git

• Use feature branches for development
• Write descriptive commit messages
• Squash commits before merging
• Use pull requests for code review
• Tag releases semantically
• NEVER commit directly to main
• NEVER force push to shared branches
• NEVER commit sensitive data

When to Use This Skill

Use this skill when you need to:

• Run SDK release onboarding for a new project
• Release packages to PyPI or GitHub
• Set up or debug CI/CD pipelines
• Configure GitHub Actions workflows
• Build multi-platform wheels
• Deploy documentation
• Coordinate multi-package releases
• Establish Git workflows
• Manage test matrices

Related Skills

• 03-nexus - Application deployment (end-user)
• 02-dataflow - Database operations
• 01-core-sdk - Runtime selection
• 17-gold-standards - Release best practices

Support

For SDK release help, invoke:

• release-specialist - Release onboarding, PyPI publishing, CI management
• release-specialist - Git workflows, releases, version management