Attribution

Source network

The publishers CHAINPRESS monitors, and what the site does with what they publish. Stories written here appear in full, crediting and linking the reporting they were built from. Everything else is an aggregated headline: the publisher’s own summary and byline, credited, with the article itself staying on their site.

What is on the site

In the feed

837

last 30 days

Written here

of those, ours

Aggregated

787

of those, linked out

Publishers · 4

Publishers monitored by Chainpress and how many of their headlines are currently showing.
Publisher	Showing
CoinDesk ↗	423 headlines
Cointelegraph ↗	371 headlines
Decrypt ↗	15 headlines
Cryptoflies ↗	0 headlines

Last updated Sep 20, 2026, 10:51 AM UTC. Counts are of headlines currently showing, not of everything ever collected.

What the site does with them

Full method →

Headlines are stored, articles are not: Each aggregated entry holds the headline, the summary the publisher put in their own feed or share card, the byline, and a link. Article text is never copied or hosted here.
Stories written here are written, not paraphrased: Where a briefing is published on a piece of news, the source reporting is read as research and a new account is written from the facts. Facts are free to use; wording is not. Source text is not retained.
A source’s photography is never reused: A credit line is not a licence. Stories written here run generated abstract plates, labelled as illustrations on the card and on the page.
Every aggregated headline links back: Those cards open the publisher’s page in a new tab, and the card names the domain it will open before you click it. Reading happens on their site, where their advertising and subscriptions are.
Collection is feed-first and rate-limited: Headlines come from publisher-operated feeds and the sitemaps listed in their robots.txt, reading only page metadata, with a delay between requests.

If you are a publisher and want out

This page used to tell publishers they could be removed “by deleting their entry from src/data/sources.ts” — an instruction to somebody with commit access to this repository, which a publisher asking to be taken down does not have. That was not a process. This is.

01Tell us what to remove: A publication name removes every headline from that outlet and stops future monitoring. A single URL removes that one item. A request about a briefing written here — one crediting your reporting — is handled the same way and is not treated differently because we wrote the words.
02No justification is needed: You do not have to explain, assert a legal position, or argue that feed syndication does not imply consent. Being the publisher is sufficient. Requests are not weighed on their merits.
03What happens then: The item or the outlet comes out of the store, the source is removed from the monitored list so it does not return, and the change goes live at the next deploy. You get a reply confirming what was removed.
04Target: Two working days to acknowledge, five to complete. If that slips the removal still happens; the acknowledgement is what is late.

The contact address is not published yetThis site has no mailbox behind it. Publishing an address that bounces would make the removal process look real while being unusable, which is worse than saying so. Until an inbox exists this process is written down and not yet reachable — an open dependency on the site owner, recorded here rather than hidden. It becomes live, along with corrections and the privacy routes, from the single NEXT_PUBLIC_CONTACT_EMAIL setting.