Use this skill to turn a recurring product-operations problem into a concrete, reviewable artifact instead of generic advice.
Workflow
- Inventory every skill folder, frontmatter description, references, evals, examples, agents metadata, and install path.
- Read
references/catalog-generation-patterns.md. - Normalize skills by user job, lifecycle area, product type, and marketplace group instead of folder order alone.
- Score quality with observable evidence: trigger clarity, reference depth, eval coverage, behavior demo, installability, and freshness.
- Produce a catalog update, gap matrix, publish checklist, and automation hooks.
When not to use
- Do not use for generic advice the base model already handles without this skill's specific artifact contract.
Guardrails
- Do not fabricate quality signals; mark missing evidence as a gap.
- Separate generated registry facts from editorial marketplace claims.
- Keep private/internal skills and public catalog content separated.
- Prefer stable skill names, links, and install commands over marketing-heavy copy.
Output format
Catalog source:
Skill count / install source:
Grouping model:
- <group> -> <skills>
Quality score table:
| Skill | Trigger | Depth | Eval | Demo | Install | Gaps |
| --- | --- | --- | --- | --- | --- | --- |
Coverage gaps:
- <gap> -> <next skill or reference>
Publish actions:
- <registry/catalog/docs/verification step>