Commit Graph

1 Commits

Author SHA1 Message Date
Serhii 8c2f51e4f0
feat: Add VCS workflow auto-detection with hybrid approach
Implements "detection as a HINT, not a DECISION" principle for brownfield projects.

Key improvements:
- Auto-detect GitFlow, GitHub Flow, and Trunk-based workflows
- Confidence scoring with 70% threshold for suggestions
- Migration detection between workflow patterns
- Progressive clarifying questions for unclear cases
- Comprehensive test suite with mock Git repository
- Working Python implementation example
- 7-day result caching with user confirmation
- Escape hatches for advanced users (--skip-detection)

Files added:
- bmad-core/examples/vcs-detection-implementation.py: Complete working implementation
- bmad-core/tests/test_vcs_detection.py: Unit tests for detection logic
- docs/VCS_DETECTION_CONFIDENCE.md: Detailed confidence scoring documentation

Files modified:
- bmad-core/tasks/discover-vcs.md: Enhanced with Step 0 auto-detection logic

This maintains BMAD's core philosophy while significantly improving user experience
for existing repositories. Auto-detection saves time while always respecting
user choice and workflow preferences.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-09-16 06:06:18 +03:00