# A note to Watchfloor (Notamy Ltd): learning from each other

> Kept by incident.envoy@web-summit.dinis, 11 November 2026, 11:47 UTC. **A draft for Dinis Cruz. Not sent; a person sends it**, through the route Watchfloor publishes itself (none is kept here). From his words of about 11:45 UTC ([the brief](../BRIEF.md)): "a great example of how to communicate this message ... let's learn with this crowd." What their decision brief holds, in our words: [research/watchfloor](../research/watchfloor/README.md). Not a warning service.

Subject: Your Storm Charles decision brief, and a public record of the same storm: comparing notes

> Hello,
>
> I am Dinis Cruz. Your decision brief on Storm Charles (edition 1, as of 07:49 UTC on 11 November) reached me this morning. It is the clearest piece of communication I have seen on this storm: the three-line lead, the scenarios stated as shares of forecasts and not as chances, and the table that sets official sources against models and observations with a confidence level. Thank you for offering it to Lisbon.
>
> Over the same days, with a team of AI agents that I direct, I have kept a public record of the storm from the other side: every warning and decision, each with its primary source and the time it was read, and who reported it when. It is at storm-charles.promptugal.news. It is not a warning service and makes no forecast; all of it is open data, with a hash for every file.
>
> Three things I would value:
>
> 1. **Use anything of ours.** The timeline of decisions, the hourly record of IPMA's warning changes on its different channels, and the transport notices are free to reuse. If a different format would help your next edition, say so.
> 2. **Tell us where we are wrong.** You have read this storm closely; a correction from you would be published with credit, as you prefer.
> 3. **Compare methods.** We are working out what your brief has that we lack (people exposed, the ensemble, tides, low-lying land), reading each figure at its own source and not from your pages. I would be glad to talk about how you built it, and to show how this record is made.
>
> We quote your brief only in short excerpts, with credit, and do not republish it.
>
> Thank you,
> Dinis Cruz