Show HN: FlowTracker – Track data flowing through Java programs

FlowTracker is a Java agent designed to monitor data flow within Java applications, providing insights into how data is read, manipulated, and written. Its primary function is to connect inputs and outputs, helping users trace the origins of output data. A demonstration of FlowTracker can be seen using the Spring PetClinic application, where it tracks HTTP requests and the generation of HTML pages based on templates and database interactions. Currently, FlowTracker is a proof of concept and may not be suitable for production due to potential performance overhead. Users can utilize FlowTracker by downloading the agent JAR and adding it to the Java command line with specific JVM options. The agent operates by modifying bytecode at runtime to maintain a mapping of in-memory data to its source, employing method interception and dataflow analysis to track local variables and method arguments. Key components include the Tracker, which holds information about tracked objects, and the TrackerRepository, a global map linking objects to their trackers. However, FlowTracker has limitations, as it does not track all values due to performance concerns and primarily focuses on textual and binary data, with limited support for numerical values. The GitHub repository provides detailed documentation on usage and implementation.

- FlowTracker tracks data flow in Java programs, connecting inputs and outputs.

- It can be demonstrated using the Spring PetClinic application.

- The agent modifies bytecode at runtime to track data origins.

- It is currently a proof of concept and may not be suitable for production.

- Limitations include a focus on textual and binary data, with some performance constraints.