What problem does it solve? Current-state data flows between systems are often described only in prose or interviews, so contradictions stay invisible and diagrams that do get drawn may not compile, omit payload and cadence, or assert flows nobody evidenced. This Skill turns flow findings into Mermaid diagrams that compile, state payload, cadence and reliability, and name their Unknowns explicitly. ## Core Features & Use Cases - Deterministic validation gate: A bundled script compiles every Mermaid diagram with mermaid-cli and runs structural lints (unlabelled edges, missing classDef, missing cadence, orphan nodes, oversized graphs), exiting non-zero on errors. - Reliability modeling: Four node classes (reliable, fragile, broken, isolated) make the maturity of each store and flow part of the diagram itself, with an explicit Unknowns list for what evidence cannot establish. - Use Case: Given interview notes saying "CRM syncs members to Policy Admin nightly and drops date of birth", produce .gener8v/flows/claims.md with a labelled, classed Mermaid diagram, a reliability table, and an Unknowns section, then iterate until the validator exits 0. ## Quick Start Map how claims data moves between our systems into validated Mermaid flow diagrams using the flow-mapping skill.
