This skill helps an agent inspect names before changing them.
Use it when code, plans, issues, schemas, states, or docs feel harder to reason about than they should. The skill looks for overloaded terms, one concept with many names, one name carrying several concepts, weak contrast, and vocabulary that will confuse future agents.
Good names make a system easier to hold in your head. Bad names create quiet tax: every reader pays a little extra, and agents often pay it badly.
Good fits:
- Reviewing a plan before tickets are written
- Cleaning up domain vocabulary
- Naming states, records, APIs, and modules
- Finding places where two concepts got merged
- Deciding whether a rename is worth it
The skill should not rename first. It should map meaning, boundaries, and tradeoffs, then suggest the smallest useful change.
# skills.sh CLI — Claude Code, Codex, Cursor, OpenCode, and more
npx skills add ratacat/ratacats-skills --skill name-review
# or the Claude plugin marketplace
/plugin marketplace add ratacat/ratacats-skills
/plugin install name-review@ratacats-skills