search_knowledge (query-based retrieval) and list_sources (permission-scoped enumeration) are both discovery-oriented and could momentarily be confused, but the descriptions clearly separate query search from browsing the allowed corpus. get_document is unambiguously a fetch-by-id operation with a distinct input signature.

All three tools follow the same verb_noun snake_case pattern (search_knowledge, get_document, list_sources) with consistent, predictable verb choices for query/fetch/enumerate semantics.

Three tools is at the low end but each maps to a distinct retrieval need (discover by query, enumerate permitted corpus, fetch full text), so there is no padding. Slightly thin for a document platform, but well-scoped for a read-only knowledge endpoint.

The discovery-to-retrieval lifecycle is covered: list what you can see, search it, then fetch full text by doc_id. Minor gaps exist, such as no way to retrieve multiple documents or inspect version history/related documents, but core workflows are not blocked.