Skip to content

when-to-use-what.md

Latest commit

 

History

History
255 lines (199 loc) · 6.07 KB

when-to-use-what.md

File metadata and controls

255 lines (199 loc) · 6.07 KB

When to Use SAI vs SAIGEN

Understanding when to use each tool will help you choose the right installation and workflow for your needs.

SAI - Software Action Interface

What It Does

SAI executes software management actions using provider-based configurations and saidata files.

When to Use SAI

✅ Use SAI when you need to:

  • Install, configure, or manage software on systems
  • Execute automated deployment workflows
  • Run software management actions in CI/CD pipelines
  • Apply consistent software configurations across environments
  • Manage software lifecycle (install, update, remove)
  • Work with existing saidata from the saidata repository

📦 Installation:

pip install sai

SAI Use Cases

  1. Production Deployments

    sai install nginx
sai configure postgresql --config prod.yaml

  2. CI/CD Pipelines

    # .github/workflows/deploy.yml
- name: Install dependencies
  run: sai install --from requirements.sai.yaml

  3. Infrastructure Automation

    sai apply --config infrastructure.yaml --dry-run
sai apply --config infrastructure.yaml

  4. System Administration

    sai list installed
sai update --all
sai remove obsolete-package

SAI Characteristics

  • Lightweight: Minimal dependencies
  • Fast: Quick installation and execution
  • Stable: Production-ready
  • Consumer: Uses existing saidata
  • Runtime: Designed for execution environments

SAIGEN - SAI Data Generation

What It Does

SAIGEN generates, validates, and manages software metadata (saidata) files using AI and repository data.

When to Use SAIGEN

✅ Use SAIGEN when you need to:

  • Create new saidata files for software packages
  • Generate metadata from package repositories
  • Validate and test saidata files
  • Update existing saidata with new information
  • Build software catalogs and inventories
  • Contribute to the saidata repository

📦 Installation:

pip install saigen
# Or with AI features
pip install saigen[llm,rag]

SAIGEN Use Cases

  1. Creating New Saidata

    saigen generate nginx --provider apt
saigen generate docker --provider brew

  2. Batch Generation

    saigen batch generate --from package-list.txt
saigen batch generate --provider apt --category web-servers

  3. Validation and Testing

    saigen validate nginx.yaml
saigen test nginx.yaml --mcp-server test-server

  4. Repository Management

    saigen repo update apt
saigen repo cache --all
saigen repo search nginx

  5. Metadata Updates

    saigen update nginx.yaml --refresh-urls
saigen update --all --check-versions

SAIGEN Characteristics

  • Feature-rich: Comprehensive generation capabilities
  • AI-powered: Uses LLMs for intelligent generation
  • Developer-focused: Built for metadata creation
  • Producer: Creates new saidata
  • Development: Designed for authoring environments

Combined Usage: SAI with Generation Support

When to Use Both

✅ Use sai[generation] when you need to:

  • Develop and test saidata in the same environment
  • Iterate quickly between generation and execution
  • Validate saidata by actually running it
  • Work on saidata contributions with immediate testing

📦 Installation:

pip install sai[generation]

Combined Workflow Example

# 1. Generate new saidata
saigen generate myapp --provider apt

# 2. Validate the generated file
saigen validate myapp.yaml

# 3. Test execution with SAI
sai install myapp --dry-run

# 4. Refine and regenerate if needed
saigen update myapp.yaml --add-action configure

# 5. Test again
sai install myapp --dry-run

# 6. Execute for real
sai install myapp

Decision Matrix

Scenario Tool Installation
Deploy software in production SAI pip install sai
CI/CD pipeline execution SAI pip install sai
Create new saidata files SAIGEN pip install saigen
Validate saidata SAIGEN pip install saigen
Contribute to saidata repo SAIGEN pip install saigen[llm]
Local development & testing Both pip install sai[generation]
Full development environment Both pip install saigen[all]
System administration SAI pip install sai
Metadata management SAIGEN pip install saigen

Typical User Profiles

DevOps Engineer (Production)

Needs: Execute software deployments reliably Uses: SAI only

pip install sai

Metadata Contributor

Needs: Create and validate saidata files Uses: SAIGEN with AI features

pip install saigen[llm,rag]

Full-Stack Developer

Needs: Both execution and generation for testing Uses: SAI with generation support

pip install sai[generation]

System Administrator

Needs: Manage software across systems Uses: SAI only

pip install sai

Platform Engineer

Needs: Build software catalogs and automation Uses: Both tools separately

pip install sai saigen

Key Differences Summary

Aspect SAI SAIGEN
Purpose Execute actions Generate metadata
Role Consumer Producer
Environment Production Development
Dependencies Minimal Comprehensive
Speed Fast Moderate
AI Features No Yes (optional)
Primary Users Operators Contributors
Output System changes YAML files

Still Not Sure?

Start with SAI if:

  • You just want to use existing saidata
  • You're deploying to production
  • You need minimal dependencies

Start with SAIGEN if:

  • You want to create new saidata
  • You're contributing to the project
  • You need metadata generation

Use both if:

  • You're developing saidata
  • You need the complete toolkit
  • You want maximum flexibility

For more help, see: