.png&w=256&q=75)
🤓 Latest Submissions
DriftLens — Repository Drift Detector
Modern software repositories often contain configuration information duplicated across README files, package manifests, lockfiles, Dockerfiles, CI workflows, and environment templates. These sources can gradually become inconsistent—for example, different Node.js versions, conflicting package managers, mismatched installation commands, or environment variables used in deployment but missing from templates. These contradictions can cause failed builds, inconsistent development environments, and difficult-to-debug deployment problems. DriftLens is an AI-assisted repository configuration drift detection tool designed to identify these contradictions automatically and explain exactly where they occur. A developer provides a local repository path through the web interface. The backend scans supported project files, extracts structured configuration facts, and passes those facts through deterministic contradiction-detection rules. Findings include severity, explanation, source file, configuration key, value, and supporting evidence. The current implementation detects Node.js version mismatches, package-manager conflicts, missing environment variables from templates, installation-command mismatches, and lockfile/package-manager mismatches. The architecture separates scanning, parsing, detection, classification, and AI-assisted repair planning, making the system modular and independently testable. The frontend provides an interactive dashboard where developers can analyze a repository, filter findings by severity, expand individual findings to inspect their evidence, and review an optional AI-generated repair plan. The AI component is not responsible for deciding whether a contradiction exists; deterministic rules perform the actual detection, while the LLM is used only to turn confirmed findings into practical repair guidance. If an LLM key is unavailable or the AI service fails, DriftLens safely falls back without breaking the core analysis pipeline. The target users are software developers, students, engineering teams, and maintainers working with repositories containing multiple configuration sources. DriftLens aims to make configuration consistency checks easier to understand and integrate into everyday development workflows. The project also includes a deliberate demo fixture containing configuration drift and an automated backend test suite covering parsers, detection rules, API behavior, and AI repair-plan behavior. The latest validation contains 167 passing tests across 15 test files.
27 Sep 2026
