Help Your Coding Assistant Find the Right Code in Large Repos

Saurabh Rob Dahal, a Developer Advocate at AWS, has released a practical workshop designed to help developers improve how they interact with AI coding assistants when navigating large codebases. The guide focuses on investigating a specific TypeScript retry bug, teaching users how to move beyond simple prompts. By utilizing tools like ripgrep for text search, ast-grep for structural analysis, and standard symbol navigation, developers can provide their AI assistants with precise evidence and context. The workshop emphasizes a structured eight-step investigation process—ranging from reproduction and mapping to verification—that ensures AI-generated fixes are accurate and verifiable. Rather than relying on the first plausible suggestion, this methodology encourages developers to trace values and understand the underlying logic, ultimately making the coding assistant a more effective partner in debugging and maintaining complex software projects.