distilllisted

Turn what we just did into a reusable skill named **`$ARGUMENTS`** (kebab-case; if missing, propose one). Follow the Anthropic skill-creator pattern: 1. **Scope it**: in one sentence, what repeatable workflow does this skill capture? If the work wasn't actually repeatable/reusable, say so and stop (don't create noise). 2. **Create** `.claude/skills/$ARGUMENTS/SKILL.md` with YAML frontmatter: - `name: $ARGUMENTS` - `description:` a precise, trigger-friendly summary starting with what it does and the concrete situations it should activate in (this is what the model matches on — be specific). - add `user-invocable: true` if it's also useful as a `/`command. 3. **Body**: write the procedure as numbered, concrete steps — the exact commands, file patterns, gotchas, and a short example. Keep it under ~150 lines; put bulk reference in a sibling file (e.g. `reference.md`) and link it (progressive disclosure). 4. If the skill needs a helper script, add it under `.claude/skills/$ARGUMENTS/` and reference it with a path **relative to the skill directory** (e.g. `./helper.py`). Do **not** use `${CLAUDE_PLUGIN_ROOT}` here — that variable is only defined when a skill ships inside an installed plugin, not for a project-scope skill living in this repo. 5. **Validate**: re-read the frontmatter — does the description clearly say _when_ to use it? Tighten if vague. 6. If this repo is a git repo, commit: `feat(skill): add $ARGUMENTS skill`. 7. Tell the user th