Set rigor levels when Kiro builds and Claude reviews

Goal: Direct two agents without over-engineering throwaway code.
Setup: agents used in source: Claude, Kiro; tools/MCPs stated: AWS Bedrock AgentCore, Strands.
Steps:
1. Split roles: Kiro builds, Claude reviews against your rules, you validate and decide.
2. Tag throwaway code explicitly, and tell both agents what the tag means; they won’t watch it for you.
3. Compliance gate: ~10 minutes before implementing, check you’re allowed to build it the way you intend.
4. Hand-check: ~20 minutes after the automated suite, test by hand.
5. Rigor budget: full rigor for code that survives, working rigor for features, spike rigor for throwaways.
Rules: ask me before anything irreversible (sending, posting, paying, deleting); don’t invent data, results or tools I haven’t given you; tell me what you couldn’t verify.
Based on: Earl Grey — https://dev.to/earlgreyhot1701d/block-zero-oh-no-claude-kiro-and-i-over-engineered-the-throwaway-5d42

Steps are from the source. Includes a “Based on” credit line.