run.yml

name: job-watcher

on:
  schedule:
    # All times in UTC. Default below is PDT (UTC-7) for 08:00 and 17:00 local.
    # Adjust if you're in a different timezone, or if you want PST (UTC-8) in winter.
    - cron: "0 15 * * *"   # 08:00 PDT
    - cron: "0 0 * * *"    # 17:00 PDT (= 00:00 UTC the next day)
  workflow_dispatch:
    inputs:
      dry_run:
        description: "Skip sending email (digest written to artifact instead)"
        required: false
        default: "false"

permissions:
  contents: write

concurrency:
  group: job-watcher
  cancel-in-progress: false

jobs:
  run:
    runs-on: ubuntu-latest
    steps:
      - name: Checkout
        uses: actions/checkout@v4
        with:
          persist-credentials: true
          fetch-depth: 1

      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
          cache: pip

      - name: Install
        run: pip install -e .

      - name: Run agent
        env:
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
          RESEND_API_KEY: ${{ secrets.RESEND_API_KEY }}
          EMAIL_TO: ${{ secrets.EMAIL_TO }}
          EMAIL_FROM: ${{ secrets.EMAIL_FROM }}
          GOOGLE_SHEETS_SHEET_ID: ${{ secrets.GOOGLE_SHEETS_SHEET_ID }}
          GOOGLE_SHEETS_CREDS_JSON: ${{ secrets.GOOGLE_SHEETS_CREDS_JSON }}
        run: |
          if [ "${{ github.event.inputs.dry_run }}" = "true" ]; then
            python -m src.main --dry-run
          else
            python -m src.main
          fi

      - name: Upload dry-run digest (if present)
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: digest-${{ github.run_id }}
          path: data/last_digest.html
          if-no-files-found: ignore
          retention-days: 14

      - name: Persist dedup DB
        if: github.event_name == 'schedule' || github.event.inputs.dry_run != 'true'
        run: |
          if [ ! -f data/seen_jobs.db ]; then
            echo "No DB to persist."; exit 0
          fi
          git config user.name "job-watcher-bot"
          git config user.email "job-watcher-bot@users.noreply.github.com"
          git add data/seen_jobs.db
          if git diff --cached --quiet; then
            echo "No DB changes to commit."
          else
            git commit -m "chore(data): update seen_jobs.db [skip ci]"
            # Retry-push in case another scheduled run touched the branch in parallel.
            for i in 1 2 3; do
              git pull --rebase --autostash && git push && break
              echo "push attempt $i failed; sleeping"
              sleep $((i * 5))
            done
          fi