Requirements analysis: catching requirement bugs before they become code

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

The problem AI-assisted engineering amplifies: "The prompt you give to the agent is the de-facto requirement now. Every vague prompt produces a vague spec or plan, and the AI agent implementing that spec produces code full of undisclosed decisions made on your behalf, without your awareness or agreement."

kiro.dev →

Connections

Challenges

What's Easy Now? What's Hard Now? - Marc's Blog Requirements analysis shows that even as coding agents get more capable, ambiguous prompts still produce requirement bugs, complicating optimistic claims about what's 'easy now' for agents.

Supports

Formal Methods at Jane Street: Agentic Coding Changes the Calculus Catching requirement bugs before implementation complements formal-methods verification of agentic code at Jane Street.

Supported by

8 Myths of Software Development with AI ACM Queue frequently covers software engineering methodology; requirements analysis as a discipline aligns with the publication's emphasis on practical, rigorous software development practice
It's time to be right. - Marc's Blog The argument that it's 'time to be right' about decisions reinforces the case for catching requirement bugs early before they propagate into AI-generated code.

Type:	link
First published:	Wednesday, 27 May 2026 at 12:00 UTC
Location:	https://kiro.dev/blog/deep-spec-analysis/
URL:	https://mattwood.fyi/i/fc959e45