What problem does it solve? Data pipeline projects often fail because business context, table relationships, and acceptance criteria are never captured in a structured, version-controlled form. This Skill guides an interactive requirements-gathering conversation and writes the results into a standardized spec.md under .modscape/changes/<name>/. ## Core Features & Use Cases - Interactive Requirements Elicitation: Collects pipeline title, goal, stakeholders, data sources, table relationships, acceptance criteria, and target tool (dbt, SQLMesh, Spark SQL, plain SQL) through structured conversation. - Business Context Capture: Probes for tacit knowledge about how data is generated, domain-specific rules, and edge cases that cannot be derived from schemas alone. - Traceability Artifacts: Generates spec.md, glossary.md, and questions.md with answered, assumed, and open questions, plus PII-safe investigation queries for ambiguous metrics. - Use Case: A data engineer starting a new dbt pipeline uses this Skill to interview stakeholders, document join keys and cardinality between source tables, and produce a spec.md ready for the design phase. ## Quick Start Ask the AI to gather requirements for a new data pipeline using @modscape-spec-requirements and answer its questions about sources, goals, and acceptance criteria.
