Impact Analysis

Before a single line of code is written, the Analyst agent scans your project to understand the blast radius of a change. It identifies affected files, downstream dependencies, test suites that need updating, and potential conflicts with in-flight work. This upfront analysis prevents the cascading breakage that often comes from AI coding without context.

How it works

The Analyst agent reads your project structure, existing tests, and architectural documentation. It traces dependencies from the components you want to change, identifies related configuration files, and flags any areas with known complexity. The output is a structured impact report listing affected files, risk assessment, and recommended approach. This report becomes the input for the Architect agent in the next phase.

Impact analysis report showing affected files and dependency graph

Screenshot coming soon

Why it matters

Most AI coding failures happen because the AI did not understand the broader context. It might refactor a function without realizing three other modules depend on its exact signature. Impact analysis catches these issues before they become bugs. It also helps estimate effort -- if a 'simple change' turns out to affect 15 files, you know upfront rather than discovering it mid-implementation.

Dependency tree visualization highlighting blast radius

Screenshot coming soon