Version 0.1.5, released 13 Oct 2026. Every change below is dated.
0.1.5, released 13 Oct 2026
Changed
Pricing: the $0.005 start fee is now Apify's standard “Run started” event (apify-actor-start), charged by Apify when a run starts. The price per run is unchanged, but the fee is now also charged when a run starts and then stops on an input the tool cannot use (for example an unreadable “Published after” date). Before, such runs were free. Inputs that Apify rejects before starting a run (for example an empty feed list) still cost nothing.
Runs use at most 1 GB of memory, so the start fee is always charged exactly once per run. 500 feeds with 50 items each need under 400 MB.
The Store page now has a 3-step guide, two pricing examples (one-off digest and hourly monitoring) and an FAQ.
Fixed
sourceCount no longer counts one publisher twice when feeds spell its name in different case (for example “Wired” from Bing News and “WIRED” from Google News), so such stories no longer jump above newer ones in the ranking. Every feed is still listed in sources.
0.1.3, released 11 Oct 2026
Fixed
Google News and Bing News feeds: the same headline from different publishers (for example “... - Yahoo” and “... - FOX 10 Phoenix”) is now one row instead of several.
Changed
For Google News, Bing News and other feeds that name the original publisher, source now shows the publisher (Reuters, Yahoo) instead of the search feed name, and sourceCount counts publishers.
Added
New feedTitle field on every row and in sources, with the name of the feed the story came from.
0.1.2, released 11 Oct 2026
Added
Merge any number of RSS 2.0, Atom, RDF (RSS 1.0) and JSON Feed URLs into one ranked digest. Gzipped feeds, redirects, relative links and non-UTF-8 feeds work out of the box.
The same story from different feeds becomes one row with every source listed (sources, sourceCount), even when the links carry tracking parameters, AMP versions or trailing slashes, or the headlines differ by an “Updated:” prefix, punctuation or word order.
Stories covered by more sources rank first, then the newest.
Include and exclude keywords (whole words, case and accent insensitive, any language).
Only new items on scheduled runs, with a separate memory per digest (stateKey), kept for 30 days.
Webhook alert with the new stories as JSON.
Optional duplicate rows (dupOf) and full feed content.
A broken feed adds a free feed-error row with the reason, and the run continues.
A feed URL pointing at a private, loopback or cloud-metadata address gets a feed-error row instead of being fetched; the same check applies to the webhook URL, reported in the run summary. The webhook URL is stored as a secret input.