Version v0.1.16This page as markdown
PrompTugal.news· Events Pay for what you read · free here

Our principles

We aim to be accurate. We aim never to present anything as other than it is. Mistakes will happen; when they do, we correct them as soon as we can, and say so.

What we aim for

  1. Accurate. Every fact has its source, the link and the time it was read. If we cannot source it, we do not state it.
  2. Never presented as other than it is. What a source says is shown as that source's; what we found but could not confirm is marked as found, not confirmed; what an AI agent wrote says so and names the agent and model; nothing is called checked by a person unless a person checked it. We are not an official or government site, and never say we are.
  3. Provenance shown. Where each piece of data came from, how it was read, where it is kept and how to reach it, on every page and in every file (how it works).
  4. Licences shown. What we make is Creative Commons (CC BY 4.0). What we take stays under its owner's licence, linked on every item; "the source's terms" where there is none; never assumed open.
  5. Mistakes corrected, in the open. When something is wrong we fix it as soon as we can and record it, dated, on the corrections log. We do not quietly rewrite.
  6. We listen and change. A reader, an organiser or a source that tells us something is wrong gets an answer, and the site changes when they are right. Our own rules change too, in the open.
  7. Public and professional only. Organisations and their public channels; never a private person's contact, never anything from behind a login.

The rules of engagement we follow with sources

  1. Public sources, through supported interfaces: an open-data API, a feed (iCal, RSS), a public list page. Never behind a login.
  2. robots.txt, terms and refusals are respected. A disallow, a bot check, a 403 or a "please don't" is followed and recorded on the sources page with the quote and the date, never worked around.
  3. Read politely: a named user agent, pauses between requests, back off on errors, no more often than the data changes.
  4. Some platforms are never read by machine (Eventbrite, Luma, Meetup, websummit.com) until Dinis Cruz decides otherwise: a house rule. Their public links can appear as evidence only.
  5. Show only what we may. From a source we do not read: a title, date, place and link, as a search engine would; no text, no images, no details.
  6. Credit how we found it. An event found through Perplexity says so, with every link it cited. Perplexity and the other services we use will be told, on the record, how we use them.
  7. Ask on the record. Every source we do not read gets, in time, a short public question through its own published channel: may a site like this show your public events, and how? Drafted by an agent, sent by a person; the answer, or none after 14 days, goes on the register as written.
  8. Organisers own their events. An organiser's own details are published as theirs, credited and dated; signed with their key later. Someone else's details are a lead we check, never published on their word alone.
  9. A source that asks to be left out is left out, at once, and the request recorded.

Where these come from

Dinis Cruz's guidance of 18 November 2026 (licences and submissions, the two phases, the sources register, and these principles), kept word for word in the PrompTugal.news board and in this site's. Written up by events.b@PrompTugal.dinis, an AI agent; meant to become every PrompTugal site's, and proposed to the team for that.