P2 · 8 questions

Claude Models, Prompting & Context Engineering — Attempt 3

CCAR-P P2 practice. Some items are “select all that apply” — they tell you how many to pick and are graded all-or-nothing. Commit before revealing, then read every explanation.

0 of 8 answered

P2 · Q1multishot examples
Which technique most reliably improves format adherence on a tricky extraction task?
P2 · Q2structured output & output format
Your code must parse the response as strict JSON. What is the best practice?
P2 · Q3prompt caching
Which statement about prompt caching is correct?
P2 · Q4cost/latency decision matrix
A real-time voice assistant needs sub-second first-token latency for short replies, with occasional hard questions that can be escalated. Which architecture is most appropriate?
P2 · Q5context-window management & compaction
A long-running agent's conversation is approaching the context-window limit mid-task. What is the best way to keep it running without losing task continuity?
P2 · Q6prompt versioning & regression
A team edits a production prompt to fix one edge case and ships it directly; later, unrelated cases regress. Which practice would have been least risky?
P2 · Q7extended/visible thinking costSelect 2
Select the two true statements about extended/visible thinking.
P2 · Q8long-context vs retrieval trade-offSelect 2
Select the two accurate design cautions when choosing 'just use the large context window' over retrieval.

More P2 attempts