skill-discoverylisted

# When to Use Two situations: deciding whether something that just came up repeatedly should become a skill instead of being handled once and forgotten, and figuring out which existing skill, if any, already covers a task before assuming none does or building a duplicate. # Worthiness Test Score a candidate on four factors before building. | Factor | Question | |---|---| | Frequency | Has this actually come up more than once, or is this the first time? | | Reusability | Would the same content serve a different task later, or is this specific to what's happening right now? | | Complexity | Is there enough structure here (rules, dials, checklists) to be worth formalizing, or is it a single fact? | | Stability | Is this settled enough to write down, or still actively changing session to session? | Rate each low, medium, or high. Don't build a numeric scale for a seven-skill family, that's precision the current scale doesn't need. A candidate scoring low on three or more factors is answered once and left in the conversation. A candidate scoring medium or above on at least three factors is a real skill candidate, run it through `skill-creation`. Low frequency alone doesn't disqualify a candidate that's high on the other three. A rule this stable and this complex is worth capturing the first time it's articulated cleanly, waiting for repetition just means re-deriving it from scratch next time. Low stability is the harder disqualifier: a skill built on a rule still in flux get