
Changing one component in a software system can unintentionally affect many other parts of the application. Developers often need to manually trace dependencies across files, services, APIs, databases, and tests before making a change. In unfamiliar or rapidly evolving codebases, this makes impact assessment time-consuming and increases the risk of overlooking affected components. Change Blast Radius Analyzer is a developer tool that helps identify the potential impact of a proposed code change before implementation. Developers select a component in a sample software codebase and describe the change they intend to make. The system analyzes relationships within the codebase and produces an impact report showing potentially affected components, related tests, APIs, database components, and risk areas. It also explains why a component was identified as potentially affected based on detected relationships. Our target users are software developers working with unfamiliar, complex, or interconnected codebases. Instead of manually navigating the repository to understand dependencies, developers can use the analyzer as an early planning step before modifying code. The project is designed around a realistic software-development workflow rather than simply generating code. Its core value is turning repository relationships into an actionable **“what could this change affect?”** view, helping developers focus their review and testing effort where it matters. For the prototype, we created a fictional e-commerce codebase containing frontend components, backend services, APIs, database components, and tests. This controlled environment allows us to demonstrate realistic dependency relationships and evaluate the analyzer's impact detection. The project combines IBM Bob 2.0 with a React/Vite frontend and Python/Flask backend. IBM Bob was used as a development partner to understand the repository, plan the architecture, scaffold the project, create the sample codebase, analyze code relationships, and develop the initial blast-radius analysis workflow.
27 Sep 2026