A short piece of fiction, written by API Evangelist. The reader is imagined; the post, and every number and quotation from it, is real: "Find The Policy, The Rule And The Standard Behind An API Decision," Kin Lane, November 11, 2026, https://apievangelist.com/2026/11/11/find-the-policy-the-rule-and-the-standard-behind-an-api-decision/
Marisol had been a platform engineer for six years, and she had a standing argument with her team about list endpoints. She read the post on a Wednesday afternoon with a cold coffee and a design review at four.
The opening line landed uncomfortably. Most API governance conversations start with a decision somebody has to defend, the post said, and her team's decision was the one it used as its example: every list endpoint must paginate. She had said that sentence in meetings more than once. If anyone had asked her where it was written down, she would have said somewhere in the wiki, probably.
She kept reading with some skepticism. A catalog of 617 rules, 283 policies and 432 standards sounded impressive, but a big number is not a useful one. What caught her was the method rather than the size. The author ran the same query, pagination, against four types and got back a 1, a 0, a 3 and a 1. The policy existed and said what she would say. It even had a line she found herself nodding at, that an unbounded list is "a loaded gun pointed at both your database and your consumers". Then came the zero for rules. The author had a firmly worded policy and no check behind it, and said so about their own catalog without dressing it up.
That was the part she respected, and also the part that made her wince, because she knew what her own answer would be. Her team had the sentence and the wiki page. What they did not have, as far as she could remember, was anything in the linter that failed a build.
She also pushed back a little. The post admits that search is free text, so a thin result may only mean the wrong word. She thought that cut both ways: the author's zero was convincing because they tried "limit" and then browsed with no query at all, but most people would stop after one search and either declare a gap or declare victory. She did not love that the tool made it easy to do the lazy version. Still, the post says as much, and she gave it credit for that.
One line stuck with her: a policy with no matching rule is "a promise without a test". She turned that over for a while. It was a bit neat, and she suspected real life was messier. Some of her team's rules were enforced at the gateway rather than in the linter, which the post's own policy text recommends, so a missing lint rule would not necessarily mean a missing check. But she could not say that with confidence either, and the doubt itself was the answer.
At four, in the design review, she did not announce a campaign. She opened the team's ruleset file, searched it for the words pagination, cursor, page and limit, and read what came back. Two hits, both about rate limits. Nothing that looked for a next link.
She added a line to the review notes saying the pagination policy had no matching lint rule, and put her own name beside it, with a note to check the gateway config before anyone wrote one. Then she closed the laptop and went to find the person who ran the gateway.