Kodza

The two tools split cleanly along a discovery/retrieval boundary: search_components returns slugs, get_component consumes them. Descriptions reinforce the pairing (search results explicitly feed get_component), so there is no plausible way to misselect between them.

Both names follow the same snake_case verb_noun convention (search_components, get_component) with a shared noun domain. The pattern is predictable and would extend naturally if more tools were added.

Two tools is thin against the typical 3-15 range, but this is a read-only catalog lookup surface where search + fetch is the canonical minimal pair. Each tool clearly earns its place; nothing feels padded or redundant.

The read lifecycle is covered end to end: discover by behavior/look, browse by category via empty query, then fetch full code, API, keyboard, dark-mode values, and install notes. Minor gaps exist (no explicit category-enumeration tool or bulk multi-slug fetch), but agents can work around these with an empty-query search.