Bob Debug Pipeline

Bob Debug Pipeline is an AI-assisted developer workflow built around IBM Bob 2.0 that helps developers turn a failing test into an understood, validated fix. The problem we address is a common but time-consuming debugging workflow: developers must inspect failing tests, trace the underlying implementation, identify the root cause, determine the expected behavior, create or validate a regression test, implement a fix, and verify that the change does not break existing functionality. These steps can require repeated manual investigation and context switching. Our MVP demonstrates this workflow using a concrete Jest debugging scenario. The test project contains a getUserById(999) case where the test expects null, but the implementation returns undefined. IBM Bob investigates the relevant files, analyzes the failing behavior, identifies that JavaScript Array.find() returns undefined when no matching element is found, and determines that the implementation does not satisfy the test contract. The proposed minimal fix normalizes the missing result to null. IBM Bob then implements the fix in the relevant service file and runs the focused Jest test. The final validation confirms that both tests pass. We built a React/Vite frontend connected to an Express/TypeScript backend to present the workflow as a clear developer tool. Users can see the bug scenario, root-cause analysis, regression test, generated fix, validation result, and the next GitHub workflow step in one interface. The target users are developers who regularly debug failing tests and maintain existing codebases. The creative aspect of the project is combining IBM Bob's repository-aware investigation and implementation capabilities with a structured debugging pipeline that makes each stage visible and verifiable. The prototype demonstrates a practical workflow from failing test to validated fix, while keeping the debugging process understandable and reproducible.