skill factory

The problem

Building a new Claude Code skill without checking your existing catalog leads to product cannibalization and overlapping functionality. Developers often rush to code the skill before defining the commercial tiering, security review depth, or marketplace positioning.

What it does

• Performs a multi-term catalog overlap check to prevent duplicating existing paid products.
• Defines free versus paid tier splits based on the diagnostic-fixer model.
• Routes skills to appropriate security review depths (one-pass vs. two-pass adversarial) based on actual portfolio precedent rather than just stated policy.
• Generates marketplace-ready listing copy, cross-sell notes, and catalog rows.
• Structures SKILL.md files using a proven pattern including dated-snapshot references for volatile data.

Why this beats prompting it yourself

This skill enforces a "catalog-first" discipline based on real-world portfolio incidents where narrow searches missed synonyms. It handles the commercial packaging layer that generic prompts skip, ensuring your skill is sellable and safe before you write a single line of logic.

Use cases

• Verifying if a new workflow idea is already covered by your current skill library.
• Determining which features belong in a free diagnostic tier versus a paid implementation tier.
• Categorizing a skill for security review based on its legal, PII, or code-execution risks.
• Writing original marketplace listing copy that doesn't leak internal repository structure.

Known limitations

Does not perform the technical drafting or evaluation loops; use skill-creator for that. Final pricing decisions remain a human call informed by the provided context.