Source catalog

What's live right now

Part of the public catalog. Sign in and the console lists every source with its last fetch time and how many times in a row it has failed. Ask for what's missing; if it's public, we can usually add it.

Central bank · policy & open-market operations
daily, 09:00
Exchange · listed-company filings
every 2 minutes
Earnings calendar · CN / HK / US
hourly
Crypto · major pair snapshots
every 5 minutes
Statistics bureau · monthly macro releases
on publication
Weibo · trending list
every 5 minutes
Reddit · new posts in a subreddit
every 10 minutes
Telegram · public channels
every 5 minutes
Podcasts · episode updates with chapters
hourly
Zhihu · questions and top answers
every 30 minutes
GitHub · repository releases
every 15 minutes
GitHub · trending by language
hourly
Hacker News · front page / Ask / Show
every 10 minutes
arXiv · new submissions by category
hourly
Product changelogs
hourly
Government & regulators · notices
daily
Trade press · latest articles
every 15 minutes
Independent blogs / newsletters
hourly
Tech media · full text
every 10 minutes
About cadence and reliability
The cadences above are targets. Real fetch rates throttle to whatever the origin tolerates. A source that keeps failing is flagged red and paused in the console, then backfilled after it recovers. Nothing is dropped quietly.
One schema

One field table, always

The domains differ. The fields don't.

FieldTypeNotes
idstringGlobally unique, feed_id:hash, safe as a primary key
titlestringOriginal title, trimmed of invisible characters
urlstringNormalized link with tracking parameters stripped
summarystringPlain-text summary or lead paragraph
contentstring \| nullFull-text HTML, only where republishing is permitted
published_atISO 8601Origin publish time, normalized to UTC
updated_atISO 8601Last edit at origin, null if unknown
langstringBody language, BCP-47
tagsstring[]Origin categories or keywords
hashstringContent fingerprint for cross-source dedupe
revisionintNumber of detected edits to this item
deletedboolRetracted at origin; item kept, flag set
Request a source

Not on the list

Send the site and the fields you need. If it's reachable without logging in and legal to fetch, it usually goes live within three business days. If it sits behind a login or a paywall, we can't.

Put these in the email

1. The site URL or the specific page
2. The fields you need (title / body / time / author…)
3. The refresh cadence you expect

What we won't touch

Anything behind a login or paywall, anything containing personal data, and any page whose terms explicitly prohibit crawling. No exceptions.

Pick one source and get it working

Pick sources in the console after signup. Two minutes later you have a feed URL and a token.