web_search and web_fetch have clearly distinct, complementary roles: one discovers ranked results, the other reads a single page. The descriptions explicitly frame fetch as the follow-up to search, so an agent can never confuse them.

Both tools follow the same snake_case verb_noun pattern with a shared 'web_' prefix (web_search, web_fetch). The convention is perfectly predictable.

Two tools is thin for a standalone server, though each earns its place as a search/read pair. It sits at the borderline of the rubric's 1-2 tool range rather than being clearly well-scoped.

For the stated purpose of live web access, the discover-then-read loop is fully covered with no dead ends. There is no obvious missing operation an agent would need for this domain.