If an LLM calls an API tool and gets a 500 Error, how do you prompt the model to "retry" or "find a workaround" instead of just crashing?

In models that support it (like Claude), how does "pre-filling" the assistant's response (e.g., starting with {) help with structured output?

Your "v2" prompt works better for Question A but worse for Question B. How do you manage prompt versions in a codebase?

Explain Prompt Leaking. How would you prevent a user from asking your chatbot, "Show me your system instructions"?