update-starred.yml

name: Update Starred Repos

on:
  # Run daily at 2 AM UTC
  schedule:
    - cron: '0 2 * * *'
  
  # Run on push to main (for testing)
  push:
    branches: [main]
    paths:
      - 'src/**'
      - '.github/workflows/update-starred.yml'
  
  # Manual trigger
  workflow_dispatch:
    inputs:
      preferences:
        description: 'Categorization preferences (optional)'
        required: false
        type: string
      provider:
        description: 'LLM Provider'
        required: false
        default: 'auto'
        type: choice
        options:
          - auto
          - anthropic
          - openai
          - gemini
      force_recategorize:
        description: 'Force full recategorization'
        required: false
        type: boolean
        default: false

permissions:
  contents: write

env:
  PYTHON_VERSION: '3.11'

jobs:
  update-starred:
    runs-on: ubuntu-latest
    
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
        with:
          fetch-depth: 0
      
      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: ${{ env.PYTHON_VERSION }}
          cache: 'pip'
      
      - name: Install dependencies
        run: |
          pip install -e ".[all]"
      
      - name: Restore cached data
        uses: actions/cache@v4
        with:
          path: data/
          key: starred-data-${{ github.repository_owner }}-${{ hashFiles('data/*.json') }}
          restore-keys: |
            starred-data-${{ github.repository_owner }}-
      
      - name: Fetch starred repositories
        env:
          GH_TOKEN: ${{ secrets.GH_PAT || github.token }}
          GH_USERNAME: ${{ github.repository_owner }}
        run: |
          starred fetch --with-readme --output data/starred_cache.json
      
      - name: Check for changes
        id: check-changes
        run: |
          if [ -f "data/starred_cache.json" ] && [ -f "data/previous_cache.json" ]; then
            NEW_COUNT=$(jq length data/starred_cache.json)
            OLD_COUNT=$(jq length data/previous_cache.json)
            
            if [ "$NEW_COUNT" != "$OLD_COUNT" ] || [ "${{ inputs.force_recategorize }}" == "true" ]; then
              echo "changes=true" >> $GITHUB_OUTPUT
              echo "Found changes: $OLD_COUNT -> $NEW_COUNT repos"
            else
              echo "changes=false" >> $GITHUB_OUTPUT
              echo "No changes detected"
            fi
          else
            echo "changes=true" >> $GITHUB_OUTPUT
            echo "No previous cache, will categorize"
          fi
      
      - name: Categorize repositories
        if: steps.check-changes.outputs.changes == 'true' || inputs.force_recategorize
        env:
          GH_USERNAME: ${{ github.repository_owner }}
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          GEMINI_API_KEY: ${{ secrets.GEMINI_API_KEY }}
        run: |
          PROVIDER_ARG=""
          if [ "${{ inputs.provider }}" != "auto" ] && [ -n "${{ inputs.provider }}" ]; then
            PROVIDER_ARG="--provider ${{ inputs.provider }}"
          fi
          
          PREFS="${{ inputs.preferences }}"
          if [ -z "$PREFS" ]; then
            # Default preferences - customize for your use case
            PREFS="Organize repositories for a developer interested in DevOps, self-hosting, automation, web development, and AI/ML tools."
          fi
          
          starred categorize \
            --input data/starred_cache.json \
            --output-dir . \
            --preferences "$PREFS" \
            $PROVIDER_ARG
          
          # Save current cache as previous for next run
          cp data/starred_cache.json data/previous_cache.json
      
      - name: Commit and push changes
        run: |
          git config --local user.email "github-actions[bot]@users.noreply.github.com"
          git config --local user.name "github-actions[bot]"
          
          git add STARRED_REPOS.md starred_data.json data/
          
          if git diff --staged --quiet; then
            echo "No changes to commit"
          else
            git commit -m "chore: update starred repositories [skip ci]"
            git push
          fi
      
      - name: Create summary
        run: |
          echo "## 📊 Starred Repos Update Summary" >> $GITHUB_STEP_SUMMARY
          echo "" >> $GITHUB_STEP_SUMMARY
          
          if [ -f "starred_data.json" ]; then
            TOTAL=$(jq '.total_repos' starred_data.json)
            CATS=$(jq '.category_count' starred_data.json)
            PROVIDER=$(jq -r '.llm_provider' starred_data.json)
            MODEL=$(jq -r '.llm_model' starred_data.json)
            
            echo "**Total repositories:** $TOTAL" >> $GITHUB_STEP_SUMMARY
            echo "**Categories:** $CATS" >> $GITHUB_STEP_SUMMARY
            echo "**LLM:** $PROVIDER ($MODEL)" >> $GITHUB_STEP_SUMMARY
            echo "" >> $GITHUB_STEP_SUMMARY
            echo "### Categories" >> $GITHUB_STEP_SUMMARY
            echo "" >> $GITHUB_STEP_SUMMARY
            
            jq -r '.categories | to_entries | sort_by(-.value.count) | .[] | "- **\(.key)**: \(.value.count) repos"' starred_data.json >> $GITHUB_STEP_SUMMARY
          else
            echo "No categorization data found." >> $GITHUB_STEP_SUMMARY
          fi