Judgment before catalog.

StackAI Learn is edited by Javier Requena, under the Gestedia project. We are not a hype agency or a comparator with five hundred logos: it is a practical directory of AI tools we test, use and document with editorial judgment.

Gestedia is the operating context: publishing, maintenance and review. Profiles and tutorials come from real use — not from rewriting press releases or bulk-generated lists.

The directory is small on purpose. We would rather publish an honest profile — what it does well, where it fails, how to start — than an endless list of brands. Tutorials teach a repeatable flow: question, deliverable, what you should see on screen and when to stop.

We do not chase an infinite catalog. Every profile has to survive one question: what concrete job is this for, and what is the first useful flow?

stackailearn.com ships as a static site on Cloudflare Pages. That lets us serve fast, update precisely and grow without a heavy CMS.

A tool ships when we can describe three things: (1) the concrete job it solves, (2) a repeatable first flow in five steps or fewer, and (3) where it fails and when to switch.

We do not invent star ratings or review counts. If we mention price or a plan, we verify it on the vendor's official site. If a UI step changed, we update the date or mark the step as outdated.

Tutorials are not button tours. Every step includes what you should see, common failure modes and decision rules — for example, when to stop chatting and open a source, or when a diff is too large to approve.

We compare tools by the step of the job —research with citations, draft with a brief, review a diff in the repo — not by an eternal ranking.

The verdict says when to use A, when B and when neither. If the answer depends on the brief, we write it with a concrete example, not with «it depends».