AI-Assisted Engineering
Working with coding agents deliberately — spec before code, skills vs agents, MCP, context and token discipline, and reviewing output that is plausible rather than correct.
Twenty questions on using coding agents deliberately rather than hopefully: where a spec earns its time, how Skills differ from subagents, what MCP is actually for, why a full context window costs accuracy, and how to review output that is confident and idiomatic and still wrong.
You get a score at the end, not a breakdown.
Leaderboard
…
Grading happens on the server, so this part needs a connection.
Taking the quiz needs an account, so the score is yours and can reach the leaderboard. Sign in with GitHub or Google — reading this page does not need it.
Sign in if you want your score on this board.