Documentation
Feed & curation
The Feed pulls content in from the outside world — RSS, YouTube, APIs, scrapes — so you can curate it, research it, and turn the best picks into entries.
Sources
A source is anything gluso polls for new items: RSS/Atom feeds, YouTube channels, HTTP APIs (with JSON-path mapping), website scrapes, or a single PDF/page URL. Set a poll interval, or hit Run now to fetch on demand.
Topics
Topics describe what you care about — each has a filtering prompt (and optional negative prompt). Incoming items are classified against your topics so the curated queue stays relevant. A topic can be scoped to specific sources or apply to all of them.
The curated queue
Every item moves through a simple lifecycle:
| NEW | Just ingested from a source (or added by hand). |
| CLASSIFIED | Matched against your topics. |
| RESEARCHED | Enriched with key claims, context, and a summary. |
| COMPOSED | Used to create a content entry — linked to it. |
| IGNORED / DUPLICATE | Filtered out of the queue. |
Research enriches an item with key claims, context, and a summary. You can also add items to the queue by hand with Add.
From pick to entry
Select any curated item and hit Create entry. Pick a target collection and the entry editor opens with the item's title, description, body, and research results prefilled as an AI prompt — generate the entry's fields in one click, then edit and publish as usual.
Once the entry is saved, the feed item is marked COMPOSED and linked to it, so the queue always shows what's been used.