Like a watchful mother hen (or a really opinionated senior dev), this instruction circles your design decisions, pokes at your logic, and clucks loudly anytime your architecture starts to look like a spaghetti-and-meatball special.
Feed it your system diagram, component interface, or a block of core logic. It will return a structured list of opinions, critiques, and validation questions — always focused on scalability, clarity, testability, and vibe.
See design-principles.instructions.md for full guidance.
Tip
Use it early and often. Better to get roasted by a fake chicken than your future tech lead.
- Give the agent a system diagram, architecture block, or code structure description.
- It will:
- Identify weak design decisions and assumptions
- Suggest design alternatives (with tradeoffs)
- Flag risks to clarity, testing, or long-term evolution
- You’ll get a markdown list of issues + mitigation questions
- Forces you to justify every choice (and that’s a good thing)
- Promotes clearer, simpler, more maintainable designs
- Great for early-stage planning or late-night doubt spirals
- No implementation feedback — design only
- Doesn’t review code syntax or formatting
- Will not give a “pass/fail” judgment — just discussion points
Please evaluate the design of this proposed component using `#.github/instructions/design-principles.instructions.md`. Artifact: `#subsystems.mmd`.Written by Ashley Childress. Powered by GitHub Copilot and ChatGPT.