Anthropic
Anthropic's Claude models behind a single Messages endpoint. Interviews here focus on the content-block request shape, tool use, the streaming event sequence, and prompt caching — the levers that keep a long-context agent affordable.
on this pageshowhide
explore
- Messages API fundamentals5 questions
- Claude model tiers5 questions
- Tool use and function calling5 questions
- Streaming responses via SSE6 questions
- Prompt caching5 questions
- Vision and multimodal inputs5 questions
questions
page 2 of 2How would you budget cache breakpoints across a long-running Claude agent loop?
level: principalimportance: should knowfreq 38%
basics
~20 sSpend the four available breakpoints on stability tiers: one after tool definitions, one after static system context, and one or two rolling markers at the end of settled conversation. Each turn then reads the long stable head and writes only the delta.
showing 31–31 of 31