Agent Skills: review-naming

Review one code change, or one design (an OpenSpec change, a spec draft or an architecture doc), for naming, consistency and layering, and rate it Low, Medium, or High. Looks for one concept with several names across layers (table, column, DTO, route, JSON field, UI label), one word with two meanings, enum values from mixed schemes, names that mislead, item kinds or owner levels that break the shared resource pattern, singular and plural folder pairs, and layering faults such as pass-through layers, one decision in several places, rules that only one entry point checks, and doubtful public surface. Each finding cites the rule it breaks. Use when asked to "review the naming", "check this spec for consistent names", "is this PR consistent with our conventions", "review this PR that changes a contract", or when a review loop runs the naming lens. Read-only; it reports findings and does not fix them.

ID: agenta-ai/agenta/review-naming

Repository

Agenta-AILicense: NOASSERTION
4,879724

Install this agent skill to your local

npx skills add agenta-ai/agenta --skill review-naming

Skill Files

The files that make up review-naming.

Download Skill

More agent skills