Guardrails Before AI Generated Code
Before we let AI start building features, we'll set up guardrails in our boilerplate. A guardrail is anything that helps us steer AI in the direction of what we need it to stick to. Or put another way, stop it from going overboard.
What kind of guardrails can we add at this point?
- Testing framework
- Linting
- Static analysis
We could go further at this point, but these cover a lot of the specific conventions we want to adhere to.