Your prompts are your product logic. But most teams treat them like this:
// This is how 90% of teams manage prompts
const response = await openai.chat.completions.create({
model: "gpt-4o-mini",
messages: [
{ role: "system", content: "You are a helpful assistant. Be concise." }, // <-- WHO APPROVED THIS?
{ role: "user", content: userInput }, // <-- WHEN DID IT CHANGE?
], // <-- WHAT VERSION IS IN PROD?
});
No version history. No review process. No way to roll back. No way to A/B test.