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?

Instead of one giant 2,000-word prompt, why might you split a task into smaller, sequential prompts?

How does asking the model to "Respond as a panel of three experts (a coder, a security lead, and a PM)" differ from asking it to "Respond as a Senior Engineer"?

In an API inference call, what is the functional difference between the System, User, and Assistant roles?