SHADOWBOX

SHADOWBOX addresses a common software engineering problem: code that works on a developer’s machine but fails in CI or another execution environment. These failures are difficult to diagnose because the root cause may lie not only in the source code, but also in runtime versions, operating-system differences, timezones, environment variables, dependencies, or CI configuration. SHADOWBOX is an environment-aware debugging and verification platform designed to turn this investigation into a structured, evidence-driven workflow. A developer provides a GitHub repository, branch, failure description, CI logs, and relevant environment information. SHADOWBOX then analyzes the repository through three parallel investigation streams: Environment, Code, and CI. The findings are synthesized into an evidence-backed root-cause diagnosis rather than relying on a single source of information. Once a probable cause is identified, SHADOWBOX creates a controlled reproduction environment using Docker, allowing the failure to be reproduced under deterministic conditions. This helps distinguish an actual environment-dependent defect from an issue that cannot be reliably reproduced. After the defect is fixed, SHADOWBOX runs the relevant validation again in the controlled environment. The final result provides evidence that the fix resolves the original failure without introducing regressions. The workflow therefore follows five stages: Ingest → Investigate → Diagnose → Reproduce → Verify. Our prototype demonstrates this workflow using a deterministic timezone-dependent date bug. The same UTC timestamp is interpreted differently depending on the execution timezone because local JavaScript Date getters are used. SHADOWBOX identifies the timezone mismatch, traces it to the relevant code, reproduces the failure in a UTC-controlled Node.js Docker environment, and then verifies the corrected implementation using UTC-safe date getters. The key innovation is connecting investigation and verification into one workflow. Instead of treating debugging, reproduction, and validation as separate manual activities, SHADOWBOX creates a continuous evidence trail from the initial failure to the verified resolution. This can help developers reduce context switching, make environment-dependent failures reproducible, and increase confidence that a proposed fix actually resolves the underlying issue. SHADOWBOX is intended for developers and engineering teams working with CI/CD pipelines, distributed development environments, and software systems where differences between execution environments can cause difficult-to-reproduce failures.