Error Budget Prompting: Set Failure Thresholds for AI-Generated Code

In SRE, an error budget is the acceptable amount of downtime before you stop shipping features and fix reliability. It turns "how reliable should we be?" from a vague aspiration into a concrete number. You can apply the same idea to AI-generated code. The Problem Most developers have a binary relationship with AI output: either they trust it completely (ship without review) or they don't trust it at all (rewrite everything manually). Neither is productive. What you need is a threshold — a clear line that says "this much failure is acceptable, and beyond this we change our process." The Error Budget for Prompts Define three numbers for any AI-assisted workflow: ## Error Budget: Code Generation - **Acceptable failure rate:** 20% of generated functions need manual fixes - **Warning threshold:** 30% need fixes → add more constraints to prompt - **Red line:** 50% need fixes → stop using AI for this task, write manually Tracking period: 1 week (rolling) Then actually track it . How I Track I