generate-snapshot

Run python -m analysis.snapshot --output analysis-snapshot.json from the repository root. Add the --pretty flag to also print the JSON to stdout. The output file is written to the repository root and is ignored by .gitignore.

The snapshot ranks files by priority_score, calculated as change_count multiplied by max_complexity. Files with high scores change frequently and are complex, making them prime candidates for splitting into smaller modules or adding tests.

The snapshot includes summary metrics, priority hotspots, git analysis (hotspots, temporal coupling, bus factor), complexity metrics per language, debt indicators (TODO/FIXME counts, large files, long functions), and dependency analysis including circular dependencies.

Yes, the git_analysis section reports bus factor data. Files with author_count of 1 and change_count of 3 or more indicate knowledge concentration risk, suggesting documentation or pair programming for those areas.

The analysis depends on git history depth and static complexity metrics, so it cannot judge code readability or business value. Temporal coupling and complexity scores are heuristics that require human interpretation before committing to refactoring work.