No user-submitted related servers found.
Scored across 25 tools
Most tools map to a distinct resource+action (crop CRUD, actuators, notifications, AI insights), but the reading cluster — get_latest_reading, get_readings_history, get_readings_summary, get_metric_series — has four similar-sounding retrieval tools whose boundaries (single crop vs all crops, raw vs aggregated) require reading descriptions carefully. Similarly get_crop vs get_overview vs get_fleet can be confused at a glance, though descriptions do disambiguate.
All 25 tools use consistent snake_case verb_noun form with a small predictable verb set (get_/list_/create_/update_/delete_/set_/send_/record_/mark_). Singular/plural nouns align with whether the tool returns one or many resources, reinforcing the pattern.
25 tools is at the heavy end for a single IoT/grow-management server; while the domain is broad (crops, readings, actuators, commands, AI, notifications), several tools could plausibly be folded together (e.g. get_unread_count into list_notifications, get_fleet into get_overview). It is defensible but borderline bloated.
Crop lifecycle (create/get/update/delete), readings (record, latest, history, summary, series), actuators and commands, AI insights/learning, and notifications are all covered, giving a solid end-to-end surface. Minor gaps exist: no way to cancel or retry a PENDING command, no bulk mark-all-read for notifications, and no credential rotation for REAL devices.