Chained AI review gates
Your team wants to wire AI into the everyday pull-request workflow for a TypeScript product: one tool drafts implementation from a ticket, another reviews diffs, another generates tests. A teammate proposes chaining them so the test-generator and reviewer both run on the implementer's output with no human in between, auto-merging if all three 'agree'. Critique this orchestration and propose where the human gates actually belong.
auto_merge_verdict(signals: list[str], touches_sensitive: bool) → str[["typecheck|deterministic|pass","reviewer|ai|pass"],false]out"auto_merge"[["implementer|ai|pass","reviewer|ai|pass","testgen|ai|pass"],false]out"human_review"[["typecheck|deterministic|fail","reviewer|ai|pass"],false]out"block"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.