Related Servers

Four tools (quick_check, explain_my_lab_results, start_lab_analysis, full_prism_report) all perform lab-panel interpretation and differ mainly in speed, depth and payment, so an agent must read carefully to pick the right one. check_care_gaps and prepare_for_my_appointment also take nearly identical closed-option inputs (condition, medicine classes, test recency) and both return prioritized guideline-cited checks plus a printable one-pager, which is a real boundary problem. The async pair start_lab_analysis/get_lab_analysis is cleanly split, which keeps this above the midpoint.

All names are snake_case and most are verb-first (get_, start_, list_, check_, explain_), which gives a readable baseline. However, a subset uses an owner-relative possessive (get_my_medical_records, prepare_for_my_appointment, start_my_phi_account, how_to_get_started_for_my_owner) while comparable tools drop it (explain_my_lab_results vs quick_check, check_care_gaps). Some names are bare noun phrases (sample_prism_report, full_prism_report) and one is a sentence-length phrase, so the convention is mixed.

At 14 tools the set sits comfortably inside a reasonable range for a server covering onboarding, records discovery, gap checking, and several lab-analysis tiers. The count is slightly inflated by redundancy — two near-duplicate care-gap tools and two tiers of the same instant-vs-thorough analysis — but nothing feels like filler.

The surface covers the full funnel end to end: biomarker lookup, methodology, a sample report, three analysis depths, async job polling, records-system discovery, appointment prep, gap checking, and account creation, all with a consistent next_steps convention. Minor gaps exist (no job cancellation or explicit expiry handling beyond the status enum, no tool to actually retrieve a stored record), but none is a hard dead end for the stated advisory scope.