A pipeline designed for real pages.
From the pasted URL to the served feed: detection, extraction, generation, deduplication, caching. Every step is built to work where there is no native feed.
Generic auto-detection
The heart of FluxRSS. In the main column, we look for the largest group of sibling nodes with the same tag and the same class signature. That group is your list of articles. Menus, footers and secondary boxes are set aside.
Native feed reuse
If the page already declares a feed (link rel=alternate, /feed, /rss) or if you paste a feed URL directly, we reuse it as-is — it's more reliable than scraping.
Clean field extraction
Title, link, date, image and summary are extracted for each article, then normalized into a valid RSS <item>.
Deduplication by guid
Each article has a unique identifier. On every refresh, duplicates are removed — your readers never see the same entry twice.
Scheduled refresh
Depending on your plan, from every 24 h to every 15 min. The feed is re-scraped and updated automatically.
Full-text extraction
Optionally, FluxRSS visits each article to enrich the summary with the full content.
Security by default
FluxRSS fetches URLs provided by users. An SSRF guard validates every address server-side and re-checks it on each redirect.
- Private, loopback and link-local IPs blocked
- Cloud metadata endpoint (169.254.169.254) forbidden
- Schemes limited to http / https
- Bounded timeout and response size
- Re-validation on each redirect
# Generer un flux
curl -X POST https://fluxrss.app/api/generate \
-H "Authorization: Bearer <token>" \
-d '{"url":"https://lemonde.fr/faits-divers/"}'
# Reponse
{ "id": "a1b2c3", "url": "https://fluxrss.app/feeds/a1b2c3.xml" }
# Servir le flux (RSS 2.0 ou JSON Feed)
GET https://fluxrss.app/feeds/a1b2c3.xml?limit=20
GET https://fluxrss.app/feeds/a1b2c3.jsonAn API to automate
Available on the Mega plan. Create feeds programmatically, fetch them as RSS or JSON Feed, plug in your webhooks.
See the Mega plan