chapter five

5 Product intent: Reverse-engineer the PRD

 

This chapter covers

  • The Curse of Knowledge: Why fluency in the schema hides the product you built
  • Prompting AI to reverse-engineer a PRD from the source code
  • Using AI to compress the whole product into one "It's X, but Y" sentence
  • Interrogating the product's competitive positioning with AI
  • Mining the code for features the product ships but never advertises
  • Predicting the roadmap from what's missing or left half-built

In Part 1, you built out your AI toolkit: you chatted with a single file, ran a workflow that mapped ten thousand files into a tutorial, and sent an agent to trace a flow on its own. For the rest of the book, we are going to aim those tools at a real codebase so you can answer the questions that actually matter: why the code exists, what its pieces are, how they work, and how to change them safely. Part 2 tackles that first question—why the code exists. In this chapter, we'll start with the most fundamental version of that question: what product is this code actually building?

5.1 Why experts misread the products they build

5.2 Every codebase builds for someone, even a library

5.3 Ask for the product in one "It's X, but Y" sentence

5.4 Competitive positioning: The one bet rivals can't copy

5.5 Read what the code admits that the marketing won't

5.6 Summary