Examples

All examples assume two environment variables:

export PDFX_URL='YOUR_ENDPOINT_URL'
export PDFX_TOKEN='YOUR_TOKEN'

---

Export a Page Tree

A page and every page under it, as one merged PDF, into My exports:

curl -s -X POST "$PDFX_URL" \
  -H "Authorization: Bearer $PDFX_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"pageId": "123456", "name": "Handbook"}'

Export One Page with a Template

curl -s -X POST "$PDFX_URL" \
  -H "Authorization: Bearer $PDFX_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"pageId": "123456", "includeChildren": false, "template": "Report", "name": "Quarterly report"}'

Export a Whole Space as a ZIP

curl -s -X POST "$PDFX_URL" \
  -H "Authorization: Bearer $PDFX_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"spaceId": "98765", "output": "zip", "name": "Engineering space"}'

Archive to a Page and Email the Team

Attach the PDF to an archive page and @mention two people there, so Confluence emails them a link:

curl -s -X POST "$PDFX_URL" \
  -H "Authorization: Bearer $PDFX_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{
    "pageId": "123456",
    "name": "Release notes",
    "attachTo": "234567",
    "mention": ["557058:aaaa-bbbb", "712020:cccc-dddd"]
  }'

---

GitHub Actions: Export on Every Release

Store the endpoint and token as repository secrets (PDFX_URL, PDFX_TOKEN). The release tag doubles as the requestId, so re-running the job doesn't export twice.

name: Export docs to PDF
on:
  release:
    types: [published]

jobs:
  export:
    runs-on: ubuntu-latest
    steps:
      - name: Start PDF export
        run: |
          code=$(curl -s -o response.json -w '%{http_code}' -X POST "${{ secrets.PDFX_URL }}" \
            -H "Authorization: Bearer ${{ secrets.PDFX_TOKEN }}" \
            -H 'Content-Type: application/json' \
            -d '{"pageId": "123456", "name": "Docs ${{ github.event.release.tag_name }}", "attachTo": "234567", "requestId": "release-${{ github.event.release.tag_name }}"}')
          cat response.json
          # 202: started. 409: this release was already exported.
          if [ "$code" != "202" ] && [ "$code" != "409" ]; then exit 1; fi

---

Node.js

const res = await fetch(process.env.PDFX_URL, {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${process.env.PDFX_TOKEN}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({ pageId: '123456', output: 'zip', requestId: `nightly-${new Date().toISOString().slice(0, 10)}` }),
});
const { status, message } = await res.json();
if (res.status !== 202 && status !== 'duplicate') throw new Error(`${res.status} ${status}: ${message}`);

Python

import os, requests

res = requests.post(
    os.environ["PDFX_URL"],
    headers={"Authorization": f"Bearer {os.environ['PDFX_TOKEN']}"},
    json={"spaceId": "98765", "output": "zip", "name": "Weekly backup"},
    timeout=30,
)
body = res.json()
if res.status_code != 202:
    raise RuntimeError(f"{res.status_code} {body['status']}: {body['message']}")

---

Tips

• Use a requestId for anything that might be retried — CI re-runs, cron jobs, webhook redeliveries. A repeat returns 409 duplicate and starts nothing.
• Use one token per integration, named after it, so Recent calls and revoking stay clear.
• Prefer a ZIP for big spaces. Merged PDFs stop at 500 pages.
• Check Recent calls when a 202 doesn't produce a file: it shows whether the export started, finished, or why it failed.