Skip to main content

Adding Profiles and Sources to Existing Hubs

Extend existing hubs by adding new sources or profiles to the hub configuration file.

Prerequisites​

  • Write access to the hub's configuration repository
  • Understanding of Hub Schema
  • Source repositories must be accessible to hub users (e.g., hosted in a public GitHub organization or one visible to your target audience)

Adding a Source​

1. Edit Hub Configuration​

Open the hub's YAML configuration file and add a new source to the sources array:

Add a new source to the sources array:

sources:
  # Existing sources...
  
  - id: "my-new-source"                    # Unique identifier
    type: "github"                         # Source type
    repository: "myorg/new-prompt-bundles" # Repository location
    name: "My New Source"                  # Display name
    enabled: true                          # Enable immediately
    priority: 75                           # Priority (0-100, higher = more priority)
    config:
      branch: "main"                       # Git branch
    metadata:
      description: "Additional prompt bundles for specialized workflows"
      homepage: "https://github.com/myorg/new-prompt-bundles"

3. Source Types​

Choose the appropriate source type for organization we recommend github source type for versionned package:

TypeUse CaseRequired Fields
githubGitHub repositoryrepository
awesome-copilotYAML collectionsrepository
apmAPM packagesurl

4. Priority Guidelines​

Set priority based on source importance:

  • 90-100: Critical organizational sources
  • 70-89: Important team sources
  • 50-69: Community sources
  • 10-49: Experimental sources
  • 1-9: Deprecated sources

Adding a Profile​

Profiles group bundles from multiple sources into themed collections.

1. Add Profile Entry​

Add a new profile to the profiles array:

profiles:
  # Existing profiles...
  
  - id: "data-science"                     # Unique identifier
    name: "Data Science Toolkit"          # Display name
    description: "Prompts for data analysis, ML, and visualization"
    icon: "📊"                             # Optional icon/emoji
    bundles:
      - id: "python-data"                  # Bundle from any source
        version: "latest"                  # Version or "latest"
        source: "my-new-source"            # Source ID
        required: true                     # Mandatory bundle
      - id: "jupyter-helpers"
        version: "2.1.0"
        source: "official-bundles"
        required: false                    # Optional bundle
    path:                                  # Optional: organize in UI
      - "development"
      - "specialized"

2. Bundle Requirements​

Each bundle in a profile needs:

  • id: Must match a bundle ID from one of the hub's sources
  • version: Semantic version or "latest"
  • source: Must reference a source ID defined in the hub
  • required: true for mandatory bundles, false for optional

3. Profile Organization​

Use the path array to organize profiles in the UI:

path:
  - "engineering"      # Top level
  - "backend"          # Sub-category

This creates a hierarchy: Engineering → Backend → [Profile Name]

Testing Changes​

1. Validate Configuration​

Before committing, validate your hub configuration:

# If you have the extension installed locally
Ctrl+Shift+P → "AI Primitives Hub: Import Hub" → [Your hub URL]

2. Test Source Connectivity​

Ensure new sources are accessible:

  • GitHub repos are public or you have access
  • HTTP URLs return valid bundle data
  • Local paths exist and contain valid bundles

3. Verify Bundle References​

Check that profile bundles exist in their specified sources:

  • Bundle IDs match exactly
  • Versions are available
  • Sources contain the referenced bundles

Publishing Changes​

1. Commit and Push​

git add hub.yml
git commit -m "Add data science profile and new source"
git push origin main

2. Update Hub Metadata​

Update the hub's metadata section:

metadata:
  name: "Engineering Team Hub"
  description: "Centralized prompt management for the engineering organization"
  maintainer: "Platform Team"
  updatedAt: "2025-01-15T10:30:00Z"  # Update timestamp

3. Notify Users​

Users can sync the updated hub:

  • Right-click hub in Registry Explorer → "Sync Hub"
  • Or: Ctrl+Shift+P → "AI Primitives Hub: Sync Hub"

Example: Complete Addition​

Here's a complete example adding both a source and profile:

version: "1.0.0"

metadata:
  name: "Engineering Team Hub"
  description: "Centralized prompt management for the engineering organization"
  maintainer: "Platform Team"
  updatedAt: "2025-01-15T10:30:00Z"

sources:
  # Existing sources...
  - id: "official-bundles"
    type: "github"
    repository: "myorg/prompt-bundles"
    enabled: true
    priority: 100

  # New source
  - id: "ml-prompts"
    type: "github"
    repository: "myorg/ml-prompt-collection"
    name: "ML Prompt Collection"
    enabled: true
    priority: 80
    config:
      branch: "main"
    metadata:
      description: "Machine learning and data science prompts"

profiles:
  # Existing profiles...
  
  # New profile
  - id: "ml-engineer"
    name: "ML Engineer Toolkit"
    description: "Essential prompts for machine learning engineers"
    icon: "🤖"
    bundles:
      - id: "model-training"
        version: "latest"
        source: "ml-prompts"
        required: true
      - id: "data-preprocessing"
        version: "1.5.0"
        source: "ml-prompts"
        required: true
      - id: "general-python"
        version: "latest"
        source: "official-bundles"
        required: false
    path:
      - "engineering"
      - "ml"

Troubleshooting​

Source Not Loading​

  • Verify repository exists and is accessible
  • Check branch name in config
  • Ensure repository contains valid bundles

Profile Bundles Missing​

  • Confirm bundle IDs exist in specified sources
  • Check version availability
  • Verify source is enabled and synced

Permission Issues​

  • Ensure you have write access to the hub repository
  • Check if the hub requires specific permissions for contributors

See Also​