The four tools have mostly distinct roles: status reading, performance metrics reading, diagnosis interpretation, and request-to-plan parsing. However, guia_status, guia_performance, and guia_diagnose all operate on telemetry and could be confused when an agent just wants 'the current situation.' The descriptions help, but boundaries are not perfectly sharp.

All tools share the guia_ prefix and snake_case style, giving a predictable namespace. The second component is less uniform: status and performance are noun-like read targets, diagnose is a verb, and plan_gather is a compound action. This is a minor deviation rather than a real inconsistency.

Four tools is well within the appropriate range for a focused advisory/diagnostics server. Each tool has a clear place: status, metrics, diagnosis, and planning. Nothing feels redundant or bloated.

The read-only, non-executing scope is coherent and covered by status, performance, diagnosis, and plan gathering. There are minor gaps around discovering what telemetry or plan types are available, but the surface supports the stated safe-advisory purpose without obvious dead ends.