Skip to content
CMMF SkillsShared tool guidance

Document: explain the actual model set

Explain the actual model set.

Producing stakeholder documentation, model catalogues, field dictionaries and Mermaid relationship diagrams from CMMF source. The YAML remains authoritative.

The complete repository, target audience, output path(s), desired bounded scope for a large graph, and any recorded design rationale/brief. Provide matching canonical source access.

Use cmmf-document to explain the Article, Author and SEO architecture to our content team. Save a catalogue and Mermaid diagram to docs/content-model.md. Explain linked authors versus embedded SEO, effective required/localised defaults and relevant field constraints. Include source and validation evidence, and label inferred rationale.

Resolves source, validates the graph, uses the documentation helper for an initial catalogue/diagram, and saves requested output through the host editor. It adapts language to the audience and supplements the generated overview with constraints, enum values, localisation, rich-text and editorial/governance detail from the real files. Large bounded views must say what was omitted. Invalid or incomplete graphs must not produce a complete-looking unexplained diagram.

Markdown catalogue, per-model field tables, Mermaid flowchart, model IDs/kinds, embedded/reference/rich-text-embed edges, repeatability and available ownership, plus audience-specific explanation and provenance/validation evidence. The helper’s field table has Field, Type, Required, Localised, Repeatable and Description columns. It includes effective required/localised defaults and model editorial/governance blocks where present.

The helper prints Markdown; it has no --output flag. Its plain Markdown includes namespace and declared CMMF version but does not automatically embed the complete source revision/hash/validation record. The skill must add that evidence; use document --format json to retain the source/diagnostics alongside the Markdown. The overview is not a complete field-constraint reference on its own.

You can trace diagram nodes/edges and field descriptions back to real models, distinguish embedded and linked content, and identify the generation source/date and validation result.

Compare generated output against the files, including defaults and rich-text reference embeds; check scope omissions; verify that added rationale is supplied evidence or labelled inference. Regenerate affected documentation after model changes. Do not accept diagrams that simplify away ownership or missing targets without explanation.

Diagrams are derived views. They neither define model semantics nor evaluate stored entries. Readers can use them for review; changes return to YAML and must be revalidated before regeneration.

Installation · Sources and versions · Worked journeys · Team acceptance