RSS Recent Items Collector

Read one public RSS 2.0, RSS/RDF, or Atom feed and append recent articles to the Apify run's default dataset. A compact standard-library-only Actor designed for repeated scheduled snapshots.

Who it is for

Newsroom watchers, blog readers, research teams, and automation builders who need normalized recent feed entries without feed-specific parsers, accounts, or API keys.

Input example

{
  "feedUrl": "https://feeds.bbci.co.uk/news/world/rss.xml",
  "sinceHours": 168,
  "maxItems": 20
}

Field	Required	Description
feedUrl	Yes	One public HTTP(S) RSS/Atom feed URL. The form prefills one feed.
sinceHours	No	Lookback in hours, default 168 (7 days); 0 disables date filtering.
maxItems	No	Maximum dataset items, default 20, maximum 100.

Output example

One dataset row per unique qualifying article (sample values are illustrative):

{
  "title": "Example headline",
  "url": "https://example.org/articles/123",
  "publishedAt": "2026-11-01T03:45:00Z",
  "guid": "article-123",
  "checkedAt": "2026-11-01T04:00:00Z"
}

• title: article title (up to 500 characters).
• url: HTTP(S) article URL (relative links resolved against feed URL).
• publishedAt: parsed RFC 2822/RFC 3339 date in UTC, or null if dates are not required and no parseable date exists.
• guid: publisher GUID or Atom ID, if supplied; otherwise null.
• checkedAt: UTC timestamp captured at run start.

Newest parseable dates are emitted first. Duplicate GUIDs (or URLs if no GUID exists) within a single run are removed.

Scheduling use case

Run daily with sinceHours: 24 and maxItems: 50. Export the resulting dataset to a workflow that aggregates articles and deduplicates across runs using guid or url.

Important: "Recent" means an article's published timestamp falls within the lookback window. This Actor is not a persistent new-item detector: repeated runs can produce the same articles. The Actor does not read or alter another run's dataset. When sinceHours > 0, undated or unparseable-date entries are excluded; set sinceHours: 0 if you need those.

Limits and behavior

• Only a single public feed; no website/feed discovery, authentication, login, credentials, or external API keys.
• Feed downloads limited to 1 MB, HTTP timeout 20 seconds, public IP destinations on HTTP(S) ports 80/443 only; private/local networks are rejected.
• Only RSS/Atom XML, not HTML, JSON Feed, or JavaScript-rendered content. DTD/entity-declaring feeds are rejected.
• No content extraction or article body crawling. A feed can be temporarily empty or contain malformed publication dates.
• No persistent state, notifications, or guaranteed exactly-once delivery. Source server restrictions, DNS, and rate limits apply.
• Failed runs print plain-English ERROR: messages and exit unsuccessfully, without tracebacks. A valid empty result creates no dataset rows.
• The Apify platform provides the run's own storage token automatically. No user-supplied Apify token is needed; it is used only for that run's input and default dataset, not as an external feed API key.

Run / deploy

Cloud: Upload the entire folder as one Apify Actor project, build it, and run it using the Console-generated input form. The Actor uses the standard Apify run environment (APIFY_IS_AT_HOME, APIFY_TOKEN, storage IDs) through the built-in Python urllib module; it does not install the Apify SDK.

Local: Place a JSON file at storage/key_value_stores/default/INPUT.json in this Actor folder and run:

$python src/main.py

Results appear as one JSON object per file under storage/datasets/default/. Or build and run the Docker image using an appropriate mounted storage/ directory.

Store placement / pricing suggestion

• Category: Automation (primary)
• Additional categories: News, Developer tools
• Suggested monetization: Pay per event, per Actor run (e.g. a provisional $0.005/run actor-start event), not per article, because published article counts vary and old articles may reappear across runs. This amount is only a starting hypothesis; validate platform costs and market pricing first.
• Billing status: Pricing is a recommendation only. No pay-per-event charge is emitted by this code. Configure/store-test Apify PPE charging separately before publishing as a paid Actor. Alternatively publish as pay-per-usage (platform costs only).