The official guidelines (and they are only guidelines—there is plenty of room for variation, are on the Computer Society's Annals website
Below are links to a few examples:
Try out your thought on me. Don't assume you know what an anecdote is and that you don't have one. When they hear the word "anecdote", some people think of a cute little story. Our definition of anecdote is much broader than that. I see anecdotes as stories that are not trying to be works of academic history writing. They are told by someone or possible a small team of people about some relatively narrow aspect of computing history in which author(s) participated or of which they otherwise have knowledge from research or personal observation. Anecdotes contribute to the history of something and typically are not about the detailed technology itself although the technology can be sketched to provide context. Occasionally the anecdotes department has been used to show a historical letter or lecture. Also occasionally, anecdotes have been about the process of researching and writing computing history, or about notable computing books or courses.
Below are statements by two prior anecdotes editors.
Calvin Gotlieb's introduction to the Anecdotes department in the first issue:
An anecdotes department is rather unusual in a journal, especially in one which has scholarly aspirations. It is therefore appropriate that there be some guidelines on what kind of stories are thought to be suitable for publication. Journals and their departments are eventually shaped 'by what is submitted regardless of any a priori opinions of the editors, but initially at least, I will be looking for two things.
A story should make a contribution to the recorded history of computing. This means that there ought to be some specifics in it by way of names, places, and dates. The contribution need not be large, but it might serve to reveal something about a personality, or throw enlightenment on an event.
The story should be (reasonably) brief-and be interesting. This last is, of course, a highly subjective judgment, but it reflects the determination of the editorial board that the Annals must not be allowed to become a stuffy journal.
Most anecdotes will probably be about people, but stories about computers, companies, or languages are welcome. We hope to avoid libel, but are not afraid of controversy. This should leave lots of scope. Contributions are invited.
James Tomayko who was Anecdotes editor from 1988 through 2001 said the following about the content of anecdotes:
The Anecdotes column is an opportunity for participants in the history of computing to contribute reminiscences of salient events. These stories can vary in scale from the origins of a term to first-person accounts of critical turning-points. Since the material in this column often represents personal views tempered or sometimes weakened by memory, the editor invites other opinions and evidence."
Take a look at complete list of anecdotes over the decades. Note the diversity of topics.
Again: try your ideas on me.