BBC News Homepage Top Stories

Purpose

Return the current set of top stories from the BBC News homepage — title, summary, canonical article URL, publication timestamp, section, and thumbnail — as a flat list. Read-only; never posts, comments, or signs in.

When to Use

"What is on the BBC News front page right now?"
Periodic / scheduled polling of the BBC's editorial front page (digest emails, dashboards, push alerts).
Bulk ingestion of BBC top stories into a downstream search / archive / analytics pipeline.
Any flow that would otherwise scrape bbc.co.uk/news or bbc.com/news HTML — the public RSS feed is orders of magnitude faster, returns the same editorial set, and is explicitly permitted by BBC's terms of use for metadata and RSS reuse.

Workflow

The BBC publishes the homepage editorial set as a public RSS 2.0 feed at https://feeds.bbci.co.uk/news/rss.xml. The feed is served by BBC's Belfrage edge with Cache-Control: public, max-age=2-5s and a self-declared <ttl>15</ttl> (minutes) — near-realtime. No auth, no cookies, no anti-bot stealth, no residential proxy. A plain HTTPS GET from a Vercel sandbox IP returns the same payload as a --proxies fetch (verified during iteration). Lead with the feed; the browser path is a true fallback that costs ~30× more and yields the same editorial set.

Fetch the front-page feed:
```
GET https://feeds.bbci.co.uk/news/rss.xml
Accept: application/rss+xml, text/xml
```
Returns text/xml; charset=utf-8, gzip-encoded, ~5 KB compressed. The <channel> block has <title>BBC News</title>, <description>BBC News - News Front Page</description>, <lastBuildDate>, <ttl>15</ttl>, and ~30 <item> children. There is no JSON variant — the feed is XML-only.
Parse each <item> block:
- <title> — CDATA-wrapped article headline.
- <description> — CDATA-wrapped one-line summary (the dek shown on cards).
- <link> — canonical article URL on bbc.com (note: feed origin is bbci.co.uk but article links land on bbc.com). Always carries ?at_medium=RSS&at_campaign=rss — strip these to canonicalize.
- <guid isPermaLink="false"> — {article-url}#{slot} where {slot} is the editorial position (0, 1, 3, 5, 7…) the BBC currently has the item pinned to on the front page. Use only the URL part for dedup; the #slot suffix changes between fetches.
- <pubDate> — RFC 822 timestamp (e.g. Tue, 19 May 2026 11:30:55 GMT).
- <media:thumbnail width="240" height="135" url="..."/> — low-res preview at ichef.bbci.co.uk/ace/standard/240/.... For a higher-res image, swap /240/ for /480/ or /1024/ in the URL.
Classify each item by URL path:
- /news/articles/{id} → standard news article (most items).
- /sport/{category}/articles/{id} → sport story (cross-promoted into the front-page feed; {category} ∈ football, tennis, boxing, cricket, rugby-union, …).
- /sounds/play/{programmeId} → BBC Sounds audio item (radio clip / podcast). No article body, just audio.
- /news/{id} (numeric) → legacy / standing item — most notably the permanent "BBC News app" promo (id 10628994, pubDate frozen at 2025-04-30). Filter out if you want only fresh editorial.
Dedupe by canonical URL (strip ?at_medium=RSS&at_campaign=rss and any #slot suffix on the guid). The feed routinely lists the same story twice with different headlines at different editorial slots — e.g. "Big game scorer Stewart and Curtis make Scotland World Cup squad" (#0) and "Stewart, Curtis and Gordon, 43, in Scotland World Cup squad" (#7) both point at /sport/football/articles/c4g94rpvx73o. Keep the earliest-slot version (lowest #N in guid) or whichever wording you prefer.
Sort if needed. The feed order is editorial (BBC's chosen front-page order), not chronological. Sort by pubDate descending if you want a "latest" timeline; preserve feed order if you want "what BBC has at the top of the page".
Optional — section feeds. Every section has its own RSS at the same shape. Verified 200 OK during iteration: world, uk, business, politics, health, education, entertainment_and_arts, technology. The legacy sci_tech slug now 404s — use technology instead.
```
https://feeds.bbci.co.uk/news/{section}/rss.xml
```
The ?edition=int query param selects the international edition view (also 200 OK).

Browser fallback

When the RSS endpoint is unreachable (rare — feeds.bbci.co.uk has been stable for ~20 years) or you need elements that aren't in the feed (live-blog placement, embedded video, "BBC InDepth" rail), drive a browser session against the rendered homepage:

sid=$(browse cloud sessions create --keep-alive \
  | node -e "let s='';process.stdin.on('data',c=>s+=c).on('end',()=>process.stdout.write(JSON.parse(s).id))")
export BROWSE_SESSION="$sid"
browse open "https://www.bbc.com/news" --remote
browse wait load --remote
browse wait timeout 1500 --remote   # progressive hydration of card rails
browse get markdown body --remote > /tmp/bbc.md

The markdown body has each top story as a contiguous block:

[![<image alt>](https://ichef.bbci.co.uk/news/480/...)](/news/articles/{id})
[<Title><Description><relative-time> ago<Section>](/news/articles/{id})

Parse by splitting on link blocks where the href matches ^/(news|sport)/(articles|world|business|.*)/articles/[a-z0-9]+$ and the visible link text begins with a capital letter. Snapshot (browse snapshot) is well-populated for the homepage (unlike, e.g., Craigslist's search page) — link refs for every story card surface — but the markdown extract is faster.

Stealth flags are not required. The bare session reaches the page; no Akamai or PerimeterX challenge fires.

Site-Specific Gotchas

bbc.co.uk/news 302-redirects to bbc.com/news from non-UK IPs. Verified during iteration from a Vercel US-region sandbox: browse open https://www.bbc.co.uk/news lands on https://www.bbc.com/news. The RSS feed origin (feeds.bbci.co.uk) does not redirect — it serves identical content from any region. If you need the UK-edition site rendering specifically, use a UK residential proxy via --proxies and a stealth session; otherwise, the international (.com) rendering is identical for top stories.
Feed <link> URLs always carry ?at_medium=RSS&at_campaign=rss tracking params. Strip them before dedup, persistence, or sharing — otherwise the same article appears under two URLs when you cross-reference against on-site visits.
The feed mixes content types. ~30 items per response = a mix of /news/articles/, /sport/{cat}/articles/, and /sounds/play/. If your downstream wants pure text-news only, filter on URL path.
One permanent "BBC News app" promo item. pubDate is frozen at Wed, 30 Apr 2025 14:04:28 GMT and link is /news/10628994. It's the only legacy-numeric-id item in the feed — easy to detect with ^/news/\d+$.
Duplicates with different headlines. Same article URL appears twice with different <title> and different #N slot suffix in <guid>. The #N numbers track the BBC's homepage editorial rail position (Top Stories rail = #0, Sport rail = #5/#7, etc.). Dedupe by canonical URL.
<guid> is NOT a permalink — isPermaLink="false" is explicit. It's {article-url}#{slot}, where {slot} changes between fetches.
Thumbnails are 240×135 default. Higher-res variants exist by URL-path substitution: /ace/standard/240/ → /ace/standard/480/ or /ace/standard/1024/. The cpsprodpb path segment is the BBC CPS image production bucket; do not modify it.
Feed cache TTL is short. Cache-Control: max-age=2-5s and <ttl>15</ttl> minutes. For polling, 30-60 seconds is sensible; sub-5s polling will mostly hit the same cached document.
No proxies, no stealth, no auth required. Confirmed across plain HTTPS fetch and --proxies fetch (both return 200 with identical bodies from a US Vercel sandbox). feeds.bbci.co.uk/robots.txt does not disallow /news/rss.xml — the BBC's terms of use explicitly permit metadata and RSS reuse (cite the URL in the feed <copyright> element if needed).
Legacy section name sci_tech is dead. https://feeds.bbci.co.uk/news/sci_tech/rss.xml returns 404. Use /news/technology/rss.xml (sci/tech content now flows through Technology + Health).
pubDate is RFC 822 in GMT. No timezone variation; convert to ISO 8601 if your schema demands it (e.g. 2026-05-19T11:30:55Z).

Expected Output

{
  "source": "BBC News - News Front Page",
  "feed_url": "https://feeds.bbci.co.uk/news/rss.xml",
  "last_build_date": "2026-05-19T13:48:45Z",
  "stories": [
    {
      "title": "Married at First Sight UK rape allegations serious, says government",
      "summary": "A BBC Panorama investigation revealed allegations that two women had been raped during filming.",
      "url": "https://www.bbc.com/news/articles/c62xv7n4xwdo",
      "article_id": "c62xv7n4xwdo",
      "section": "news",
      "published_at": "2026-05-19T11:30:55Z",
      "editorial_slot": 0,
      "thumbnail": {
        "url": "https://ichef.bbci.co.uk/ace/standard/240/cpsprodpb/113e/live/1244ba40-5327-11f1-b682-cf91850925ea.jpg",
        "width": 240,
        "height": 135
      }
    },
    {
      "title": "Ebola outbreak may be spreading faster than first thought, WHO doctor warns",
      "summary": "Hundreds of cases are suspected in central Africa but experts fear the actual number may be much higher.",
      "url": "https://www.bbc.com/news/articles/ceqp11gn1l8o",
      "article_id": "ceqp11gn1l8o",
      "section": "news",
      "published_at": "2026-05-19T12:24:07Z",
      "editorial_slot": 0,
      "thumbnail": {
        "url": "https://ichef.bbci.co.uk/ace/standard/240/cpsprodpb/ff64/live/547a9890-536c-11f1-89a3-d1f559421220.jpg",
        "width": 240,
        "height": 135
      }
    },
    {
      "title": "'Big game scorer' Stewart and Curtis make Scotland World Cup squad",
      "summary": "Ross Stewart and Findlay Curtis are named in Scotland's World Cup squad but there is no place for Lennon Miller.",
      "url": "https://www.bbc.com/sport/football/articles/c4g94rpvx73o",
      "article_id": "c4g94rpvx73o",
      "section": "sport/football",
      "published_at": "2026-05-19T10:03:02Z",
      "editorial_slot": 0,
      "thumbnail": {
        "url": "https://ichef.bbci.co.uk/ace/standard/240/cpsprodpb/434f/live/bc3d9850-536d-11f1-89a3-d1f559421220.png",
        "width": 240,
        "height": 135
      }
    }
  ]
}