About Dan's Brief
Dan's Brief is an automated daily briefing. Every morning it reads a fixed list of public RSS feeds, scores each article, and writes one brief that summarizes the day and connects the pieces. Nobody edits it before it is published.
How a brief is made
- Fetch. Every feed below is pulled several times a day. Items from the last 36 hours are kept and duplicates across feeds are merged.
- Triage. The title and summary of each article are scored by a language model (claude-haiku-5-5) for importance on a 1 to 5 scale, assigned a topic, and given a one-line factual summary.
- Write. Articles from the last 24 hours rated 2 or higher are handed to a second model (claude-sonnet-5-5), which writes the brief in fixed sections: top stories, one section per topic, themes and analysis, and a watch list. It is instructed to work only from the supplied material and to link every claim to its source.
Importance scale
| Score | Label | Meaning |
|---|---|---|
| 5 | Critical | Major and widely consequential: an actively exploited critical flaw, a large breach, a landmark product or regulatory event. |
| 4 | High | Clearly notable for a technical reader. |
| 3 | Notable | Worth a line in a daily digest. |
| 2 | Minor | Minor; included only on a slow day. |
| 1 | Noise | Noise: marketing, listicles, deals, rumours. |
Sources
Currently 18 feeds. Each brief links the specific articles it drew on.
Computer science
IT and infrastructure
Cyber security
Accuracy
The summaries are produced by a language model from feed titles and excerpts, not from full articles. They can misstate a number, a version, or a CVE identifier, and they can miss a story that a feed did not carry. Treat the brief as a map of what to read, and read the linked sources before acting on anything.
Subscribe
An Atom feed of every brief is at /feed.xml.