22 October 2026 · Claude Code skill usage · No frontend outcome test
A skill was used. Was it useful?
If you have several frontend skills installed: check which ones ran before adding another. In eligible Claude Code terminal sessions, /skill-doctor reports usage and listing cost. It does not grade the screen.
Claude Code documentation says /skill-doctor flags listed skills that have never been invoked. It requires version 2.1.252 or later and feature-flag fetching. The report covers session skills other than bundled and enterprise skills; Remote Control does not return it. We did not run the command.
Two different questions
| Question | Check | Limit |
|---|---|---|
| Did this skill run? | Use /skill-doctor for usage history where supported. For one task, invoke a user-invocable skill directly or inspect the session. | A past invocation does not identify the instructions used in this task. |
| Did it help this screen? | Run the same brief and project in fresh sessions, with and without one pinned skill. Check visual-system fit and working behavior separately. | A single screenshot preference is not a functional or user result. |
A Claude Code user asked when skills would help a working process and described repeated UI prompt tweaks. A saved project brief may be enough; a ready-made skill is a candidate, not a necessary upgrade. A separate open issue asks how automatic selection handles overlapping descriptions. That question remains distinct from same-name command precedence documented by Claude Code.
Evidence status: official documentation and individual questions inspected. No pool-run /skill-doctor check, multi-skill comparison, or paired frontend output test.