summarize-transcripts.yml

# Gera lições ISS a partir de .vtt em downloads/ via API Cursor (@cursor/sdk).
# Publica em content/<disciplina>/ e atualiza content/lessons.json (só aulas em falta no catálogo).
# Anexa prompts ISS + contexto downloads/documents (config/documents-context.json).
# Requer secret: CURSOR_API_KEY (Dashboard Cursor → Integrations ou service account).
# Opcional: DISCORD_WEBHOOK_URL — notificação no fim do job (sucesso ou falha).
#
# Ajuste TRANSCRIPT_SUMMARY_MAX_FILES: 0 = todos os .vtt (pode demorar e consumir quota).
# Proteção de branch em main pode bloquear o push — use bypass com PAT ou regra para o bot.

name: Resumir transcrições (Cursor SDK)

on:
  workflow_dispatch:
    inputs:
      max_files:
        description: "Máximo de ficheiros .vtt a processar (0 = todos)"
        required: false
        type: string
        default: "5"
      force:
        description: "1 = regenerar mesmo que a lição em content/ esteja atualizada"
        required: false
        type: string
        default: "0"
  schedule:
    # Diário às 09:30 (America/Sao_Paulo) — comente se não quiser execução agendada
    - cron: "30 9 * * *"
      timezone: America/Sao_Paulo

concurrency:
  group: summarize-transcripts-${{ github.repository }}
  cancel-in-progress: true

permissions:
  contents: write

jobs:
  summarize:
    runs-on: ubuntu-latest
    timeout-minutes: 120
    steps:
      - name: Checkout main
        uses: actions/checkout@v4
        with:
          ref: main
          fetch-depth: 1

      - name: Configurar Node.js
        uses: actions/setup-node@v4
        with:
          node-version: "22"

      - name: Instalar conversão Office→PDF e pdftotext
        run: |
          sudo apt-get update
          sudo apt-get install -y --no-install-recommends \
            libreoffice-impress libreoffice-writer poppler-utils

      - name: Substituir .ppt/.pptx por PDF em downloads/documents
        run: node .github/scripts/convert-office-documents.mjs

      - name: Instalar dependência Cursor SDK
        shell: bash
        run: |
          set -euo pipefail
          test -n "${{ secrets.CURSOR_API_KEY }}" || { echo "Secret CURSOR_API_KEY não definido no repositório."; exit 1; }
          npm init -y >/dev/null
          npm pkg set type=module
          npm install @cursor/sdk@latest

      - name: Gerar lições ISS em content/ a partir dos .vtt
        id: summarize
        shell: bash
        env:
          CURSOR_API_KEY: ${{ secrets.CURSOR_API_KEY }}
          CURSOR_MODEL_ID: ${{ vars.CURSOR_MODEL_ID }}
          TRANSCRIPT_SUMMARY_MAX_FILES: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.max_files || '5' }}
          TRANSCRIPT_SUMMARY_FORCE: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.force || '0' }}
          TRANSCRIPT_SUMMARY_MAX_CHARS: "100000"
          TRANSCRIPT_SUMMARY_MAX_ATTEMPTS: "3"
          DOCUMENT_CONTEXT_MAX_CHARS: "80000"
          DOCUMENT_CONTEXT_MAX_FILE_CHARS: "25000"
        run: |
          set -o pipefail
          node .github/scripts/summarize-transcripts.mjs 2>&1 | tee summarize.log
          SUMMARY="$(grep -E '^Concluído\. Publicados:' summarize.log | tail -1 || true)"
          FAILURES="$(grep -E '^Concluído\. Publicados:.*\| falhas: [1-9]' summarize.log | tail -1 || true)"
          if [[ -n "$FAILURES" ]]; then
            echo "::warning::Pipeline concluiu com falhas parciais — ver log summarize.log"
          fi
          if [[ -z "$SUMMARY" ]]; then
            SUMMARY="(resumo não encontrado no log — ver Actions)"
          fi
          echo "pipeline_summary=$SUMMARY" >> "$GITHUB_OUTPUT"
          LAST="$(grep -E '^Publicado:' summarize.log | tail -1 | sed 's/^Publicado: //' || true)"
          echo "last_published=$LAST" >> "$GITHUB_OUTPUT"
          NOTIFY="$(grep -E '^NotifyDiscord:' summarize.log | tail -1 | sed 's/^NotifyDiscord: //' || true)"
          IFS=$'\t' read -r DISC_TITLE LESSON_TITLE LESSON_DESC <<< "$NOTIFY"
          DISC_TITLE="$(echo "${DISC_TITLE:-}" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
          LESSON_TITLE="$(echo "${LESSON_TITLE:-}" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
          LESSON_DESC="$(echo "${LESSON_DESC:-}" | sed 's/^[[:space:]]*//;s/[[:space:]]*$//')"
          {
            echo "last_discipline_title=$DISC_TITLE"
            echo "last_lesson_title=$LESSON_TITLE"
            echo "last_lesson_description<<EOF"
            echo "$LESSON_DESC"
            echo "EOF"
          } >> "$GITHUB_OUTPUT"

      - name: Exportar lições → JSON estruturado
        run: node .github/scripts/export-lessons-json.mjs

      - name: Commit e push para main
        id: commit
        shell: bash
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          set -euo pipefail
          git config user.name "github-actions[bot]"
          git config user.email "41898282+github-actions[bot]@users.noreply.github.com"
          git add -- 'content/' 'config/' 'jsons/' 'agents/content-summary-agent.md' 'agents/content-summary-style-guide.md'
          git add -- 'downloads/documents/**/*.pdf' 2>/dev/null || true
          git add -u -- 'downloads/documents/' 2>/dev/null || true
          if git diff --staged --quiet; then
            echo "Nada para comitar."
            echo "committed=false" >> "$GITHUB_OUTPUT"
            echo "commit_info=Nada para comitar (content/ já atualizado)." >> "$GITHUB_OUTPUT"
            exit 0
          fi
          git commit -m "chore: lições ISS automáticas (VTT → content/) [skip ci]"
          git push origin HEAD:main
          echo "committed=true" >> "$GITHUB_OUTPUT"
          echo "commit_info=Commit e push para \`main\` concluídos." >> "$GITHUB_OUTPUT"

      - name: Notificar Discord
        if: always()
        shell: bash
        env:
          DISCORD_WEBHOOK_URL: ${{ secrets.DISCORD_WEBHOOK_URL }}
          JOB_STATUS: ${{ job.status }}
          PIPELINE_SUMMARY: ${{ steps.summarize.outputs.pipeline_summary }}
          LAST_PUBLISHED: ${{ steps.summarize.outputs.last_published }}
          LAST_DISCIPLINE_TITLE: ${{ steps.summarize.outputs.last_discipline_title }}
          LAST_LESSON_TITLE: ${{ steps.summarize.outputs.last_lesson_title }}
          LAST_LESSON_DESCRIPTION: ${{ steps.summarize.outputs.last_lesson_description }}
          COMMIT_INFO: ${{ steps.commit.outputs.commit_info }}
          MAX_FILES: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.max_files || '5' }}
          FORCE_FLAG: ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.force || '0' }}
        run: bash .github/scripts/discord-notify.sh