← Documentation: RSS → AI → Instagram Publishing
The stages below explain the supplied node chain and the data it passes forward. Use this page to locate a setting or diagnose where an execution stopped.
Schedule Trigger starts a run and RSS Read retrieves feed items. Aggregate combines title[], content[] and link[].
group the news into 1 item creates an articles array with numeric IDs and a forAI preview. Best Article finder selects up to three IDs. Find the urls the ai outputted resolves those IDs against the original articles, and All url's to 1 string prepares the links.
This preserves the source mapping. Inspect the chosen IDs and resolved links when selection is empty or incorrect.
News-writer prepares the Instagram caption from the supplied links and editorial prompt. Message a model creates a short headline. The supplied writing rules use a professional tone without Markdown or emojis.
Generate an image creates the visual. Blur_image, Add_textTitle-toimage and Add_textReadMore-toimage apply blur and overlays. Coordinates depend on the image size, so inspect the final binary after changing dimensions or text length.
The Drive upload uses your configured folder and provides the source URL for Cloudinary. Upload an asset from URL returns the media URL used by Instagram params, alongside caption and target account ID. The supplied file-cleanup node deletes its temporary Drive file.
The Gmail node sends generated content to the configured recipient; it is not an approval gate.
The Graph API nodes create a media container, check status_code and publish through media_publish when the container is FINISHED. Inspect the status and final response rather than assuming an uploaded image is already a published post.
After a publishing failure, check the account and container state before retrying so you do not create duplicate posts.