Each tool has a reasonably distinct role, but list_pages, search_pages, and browse_tree all enumerate pages and their boundaries depend on the descriptions rather than the names. The get_page vs get_page_by_id pair is only distinguished by path-vs-ID, which is clear enough once read but easy to misselect at a glance.

Every tool carries the wikijs_ prefix and a verb_noun form (search_pages, list_pages, list_tags, get_page, get_page_by_id), so the set reads predictably. Minor deviations: browse_tree uses a different verb style, and get_page/get_page_by_id are near-duplicates in name shape.

Six tools is well within the ideal range and each covers a distinct read path (search, fetch by id, fetch by path, list metadata, tree browse, tags). Nothing feels padded or redundant.

The surface is entirely read-only — there is no create, update, move, or delete page operation, so any agent asked to author or maintain wiki content will hit a dead end. Tag listing without any tag-mutation or page-write tool reinforces that this is a partial lifecycle.