Management & Export

History & Retention

All your AI-generated batches are saved in the History page.

Data Retention

Generated pin data follows a short retention policy:

• Unscheduled History pins are retained for 7 days.
• Pending scheduled pins stay visible in History and Scheduler until they publish, fail, or are cancelled, even if they are scheduled more than 7 days out.
• Stored scheduled images are removed from Cloudflare R2 24 hours after publish, failure, or cancellation.
• Completed scheduled records are removed after 7 days.

Action Required: Download unscheduled pins (ZIP or CSV) within 7 days if you want to keep the images locally. For future scheduled pins, keep them in Scheduler until Pinterest has published them.

Editing Generated Pins

You can edit any generated pin directly from the History page:

1. Click on a pin to open the edit modal.
2. Update the title, description, keywords, or destination URL.
3. Save — changes are applied immediately.

This is useful for fine-tuning AI-generated content before scheduling.

Scheduling Status Badges

Pins on the History page show scheduling status badges when they've been scheduled:

• Pending: Scheduled but not yet published.
• Published: Successfully posted to Pinterest.
• Failed: Publishing failed (check the Manage Pins tab for details).

When scheduling a batch, PinBuilds checks for already-scheduled pins and warns you to prevent duplicates, with options to skip duplicates or schedule all anyway.

Batch Merging

Single-pin batches from the same domain generated within 30 minutes are merged into a single visual entry. This keeps your History clean when generating one pin per article.

Publishing to Pinterest

There are two ways to get your pins onto Pinterest:

Direct Scheduling (Recommended)

If you've connected your Pinterest account in Settings, you can schedule pins directly from PinBuilds:

1. From History: Select a batch, click "Schedule to Pinterest", choose your boards and timing, and schedule.

PinBuilds publishes pins automatically via a background process. See the Scheduling docs for full details on configuration and managing scheduled pins.

External Scheduler Pack

If you use tools such as Postiz, Publer, Make, Zapier, or n8n:

1. Go to History and open a batch.
2. Select specific pins, or leave everything unselected to use the full loaded batch.
3. Click "Scheduler Pack".
4. Download the ZIP containing:
   • pinbuilds-scheduler-pack.json for automation tools.
   • pinbuilds-scheduler-pack.csv for spreadsheet review or manual imports.

Each scheduler pack includes title, caption, image URL, destination URL, alt text, keywords, source type (website, Shopify, Etsy), and suggested publish time.

Send to Postiz, Publer, or Webhooks

You can also send pins directly to a saved scheduler integration:

1. Go to Settings > Scheduler Integrations.
2. Add a Webhook, Postiz, or Publer integration.
3. For Postiz or Publer, use Fetch Targets to copy account/channel IDs, then save the target IDs on the integration.
4. Return to History, open a batch, and click "Send to Scheduler".

Direct Postiz/Publer sends are intentionally limited to 10 pins at a time because these APIs import image media before creating posts. For larger batches, use Scheduler Pack export or a webhook automation.

CSV Export (Manual)

If you prefer to use Pinterest's Bulk Upload tool:

1. Go to History and select a batch (or specific pins).
2. Click "Export CSV".
3. Configure your schedule (start date, pins per day).
4. Download the CSV.
5. Go to Pinterest Bulk Create and upload the file.

Article Tracker

PinBuilds permanently tracks every article URL you've processed:

• Server-side tracking: The system maintains a persistent record of all generated article URLs — this data is permanent (unlike pin history which has 7-day cleanup).
• Duplicate prevention: When browsing sitemaps, already-processed articles are clearly marked so you don't accidentally generate pins for the same article twice.
• WordPress integration: The WordPress plugin queries the server to determine which posts still need processing, using the server as the single source of truth.