CodePulse

As software projects grow, codebases become increasingly difficult to understand, maintain, and improve. Developers may encounter duplicated logic, dead or unused code, large or risky files, and maintainability issues scattered across a repository. Although traditional static-analysis tools can detect many of these problems, their results can still be difficult to interpret and may not clearly tell developers what should be reviewed first. This creates a gap between identifying code issues and understanding where developers should focus their attention. CodePulse addresses this gap by turning repository analysis into an understandable, explainable, and prioritized code-health report. Users provide a public GitHub repository link, and CodePulse processes the codebase through a structured analysis pipeline. First, the system retrieves the repository and identifies supported Python source files. It then performs deterministic static analysis to inspect the source code and detect code-health concerns, including duplicate code and potential dead or unused code. Next, the scoring layer processes the analysis results and calculates file-level risk information. This allows CodePulse to prioritize files that may deserve attention instead of requiring developers to manually inspect every file in the repository. IBM Bob provides the AI-assisted explanation layer. When a user requests an explanation for a selected file, CodePulse sends Bob relevant context including the file path, programming language, source-code context, duplicate-code findings, and dead-code findings. IBM Bob interprets this information and generates concise, developer-friendly code-health observations. Finally, CodePulse presents the analysis through an interactive web dashboard. Developers can review repository-level information, explore prioritized files, inspect individual findings, and request IBM Bob explanations for selected files. The target users are software developers, engineering teams, and students working with growing or unfamiliar Python codebases. The interaction is intentionally simple: provide a public GitHub repository, let CodePulse analyze it, explore the prioritized code-health results, and use IBM Bob when additional explanation is needed. What makes CodePulse different is the combination of deterministic static analysis, risk-based prioritization, AI-assisted explanation, and visual presentation. Rather than replacing static-analysis techniques, IBM Bob complements them by turning scanner findings and source-code context into explanations that are easier for developers to understand.