conflict-detection

Duplicate detection automatically scans your story-tree database to identify near-identical titles, overlapping acceptance criteria, and competing solutions using lightweight text similarity. Run it before planning features or periodically in CI to catch conflicts early and prevent redundant work.

Scope overlap analysis identifies stories where one subsumes or partially covers another, flagging feature creep risks. The Skill applies domain-specific synonym normalization and outputs conflict reports with confidence scores and similarity metrics to distinguish true overlaps from false positives.

Yes, conflict detection uses only Python standard library and sqlite3 for data access, requiring no external dependencies. This lightweight approach keeps setup minimal while performing text similarity analysis directly against your story-tree database.

Competing approach detection flags stories that solve the same problem with incompatible solutions. Run the Skill on request to audit existing stories, receiving structured JSON output with conflict type, confidence levels, reasons, and similarity scores for each flagged pair.

Conflict reports output in both human-readable text and structured JSON formats, including conflict_type, confidence, reason, and similarity scores. This dual format supports both quick review and downstream integration with other analysis or management tools.