Content sources

Four source types

• RSS. Any RSS-ish feed: blogs, Substack, Reddit as JSON, company changelogs.
• YouTube. A specific channel URL. We pick up new uploads and ingest the title, description, and (when available) transcript.
• Podcast. A podcast show URL (Apple, Spotify, or direct feed). New episodes become raw items.
• Webhook. A URL we generate that you POST to from any external tool — Zapier, Make, your own script.

You can also paste a single article URL into the Sources page to ingest that article one time (without subscribing to a feed). Good for one-off inspiration.

Library vs private

The Sources page has two tabs. Library sources are tenant-wide and can be assigned to any brand profile. Private sources are owned by a single brand profile and never leave it. Solo accounts only ever see the private tab — there’s nothing to share across.

Adding a source

Click Add source, pick a type, paste the URL, and confirm. For RSS we validate the feed up front; for YouTube we resolve the channel ID; for podcasts we find the RSS feed behind the show page. The first ingest runs within an hour; subsequent polls happen a few times per day.

Webhooks

Add a webhook source and we generate a secret URL. POST JSON of any shape to that URL and we store the payload as a raw item on the matching brand profile. Use it to wire in sources we don’t natively support — a CRM note, a Slack message, a headline from your monitoring tool.

Removing or deactivating

The toggle on a source row pauses ingestion without deleting anything — past raw items stay, but we stop polling. The X deletes the source and all items ingested from it. Use the toggle when you expect to come back; use the X when you won’t.