Ask HN: How would you lint natural-language specs before LLMs run them?
1 ceyhunkazel 0 9/18/2025, 10:16:24 PM
We lint code before execution—what’s the practical equivalent for natural-language specs before an LLM/agent runs them? I’m not looking for “NL ≠ code” debates; I’m after concrete designs, rules, and tooling that catch mistakes pre-execution.
No comments yet