The three writing tools split along a clear workflow: lint detects mechanical violations, rubric is a static reference of platform scoring rules, and compose generates prompt constraints. Lint vs. rubric could be momentarily confused since both are 'check my text against rules', but the descriptions frame one as automated scanning with line numbers and the other as a manual checklist. groups_list is unmistakably a meta/introspection tool.

All lowercase snake_case with an action-oriented second token, so naming is readable and predictable. The three writing_* tools share a clean domain prefix, while groups_list is a noun_verb outlier that breaks the domain_action pattern.

Four tools is lean but defensible for a focused writing-quality server: one generates, one checks, one provides reference criteria, plus a discovery tool. Nothing is redundant, though the surface is thin enough that the single meta tool feels like it belongs to a larger, uninstalled product.

The compose → lint → rubric loop is covered, but rubric is a passive reference with no tool to actually score a draft against it, and there is no update/fix operation to apply the lint suggestions. The presence of groups_list implies additional capability groups exist outside this server, so the visible surface is intentionally partial.