The DailyDossier bot

DailyDossier publishes one daily dossier per theme, synthesized from a small fixed roster of podcasts and written publications. This page says how our fetcher behaves.

Who we are

DailyDossier, a one-person publication at dailydossier.ai. Contact: hello@dailydossier.ai.

What we fetch

RSS and Atom feeds that publishers choose to offer — podcast feeds and the full-text feeds of written publications — about once an hour, from a short hand-picked roster. We do not crawl pages, follow links out of a feed, fetch paywalled content or bypass any access control. A feed item that is truncated or marked for subscribers only is not used at all.

Each request gives up after 20 seconds and stops reading at 5 MB, so a slow or very large feed costs the publisher one short connection and nothing more.

How we identify ourselves

User-Agent: DailyDossier/1.0 (+https://dailydossier.ai/fetcher). One source (a public-domain US government feed that refuses non-browser agents) is fetched with a browser User-Agent; every other request carries the string above.

What we honour

Before we add a source we read its robots.txt, Cloudflare Content-Signal (we are an ai-input use), RSL and TDM reservations and its own terms by hand, and we re-read them monthly. Any of them refusing automated or AI use is read as an opt-out and the source is dropped.

What we keep

The text of a written article is kept only while it is processed and is deleted within 24 hours of the dossier that used it. What remains is a short summary, one quoted anchor of at most fifteen words, the outlet, the author and the link to the original. Podcast audio is transcribed for processing and not republished.

How to opt out

Email hello@dailydossier.ai naming the feed. The source is removed and its stored items deleted promptly, normally within a few days; edition pages stop citing it, and emails already delivered cannot be recalled.

Automated summaries of what was said on each show — not claims by DailyDossier and not independently verified.