What Makes a Persistent Instruction Useful
A free lesson from Swipe AI · Prompt Engineering, level 2 · updated October 2026
By the Swipe team, founded by Colin · how we make lessons
A persistent instruction earns its place by covering things that should genuinely hold true across every request: a default tone, a standard output format, a language preference, or a short list of things to always avoid. These are the settings you would otherwise have to repeat in every single message.
Good candidates read the same no matter what the actual task turns out to be: 'always explain acronyms on first use,' 'never recommend a specific brand,' 'default to metric units unless asked otherwise,' 'keep responses under 200 words.' None of these depend on knowing what the next request will be about.
A useful way to test a candidate rule is to imagine it attached to ten completely unrelated future requests — one about a recipe, one about a code bug, one about a vacation itinerary. If the rule still makes sense every time, it belongs in the persistent layer. If it only makes sense for one of those ten, it belongs in that one user message instead.
Good persistent instruction candidates: - Always respond in plain, jargon-free language - Default to bullet points for anything with more than three items - Never make up a fact you are not confident about - Keep a warm but professional tone
Which of these is the best candidate for a persistent instruction?
A good persistent instruction should still make sense no matter what the next ____ turns out to be about.
A persistent instruction should generally contain rules that would make sense attached to almost any future request.
Free ai tools
Keep your streak going in Swipe AI
Swipe through every level of Prompt Engineering on your iPhone — with XP, stars, streaks, a personal advisor and the full track beyond these free levels.
Get the Swipe apps →