Over-reliance on agent scaffolding
Over a quarter, your team's velocity rises because nearly every TypeScript feature is now scaffolded by an AI agent. Then your most senior engineer goes on leave, an agent-generated state-machine module starts misbehaving in prod, and nobody on the team can explain how it works or confidently fix it — they only know how to ask the agent to 'fix it,' which makes it worse. How do you diagnose and address the over-reliance, without throwing away the productivity?
rank_comprehension_debt(module_names: list[str], explainer_counts: list[int], invariant_counts: list[int], invariant_test_counts: list[int]) → list[str][["auth_state_machine","billing_calc","email_templates"],[1,3,2],[6,4,1],[2,4,1]]out["auth_state_machine"][["a_mod","b_mod","c_mod"],[0,2,2],[1,5,3],[1,0,0]]out["a_mod","b_mod","c_mod"]Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.