Server has 6 tools with basic schemas and descriptions, but significant gaps limit production readiness. All tools have input schemas with type definitions and descriptions present, meeting baseline requirements. However, descriptions are generic (avg ~60 chars, well below the 194-char baseline for A+ tools), lack actionable context, and omit critical details like prerequisites, return structures, and error guidance. Parameter descriptions are minimal. Output schemas are undocumented, LLMs cannot plan downstream calls. No error handling guidance. Tool naming follows verb_noun convention (good), but descriptions fail to explain WHEN to use each tool vs. alternatives (e.g., list_properties vs. check_availability both query property data but serve different intents). create_reservation is the only write tool but lacks confirmation/dry-run pattern for irreversible operations. No pagination guidance despite list tools accepting limit/skip. Overall: functional but below production baseline.
Check availability of properties for specific dates
Create a new reservation in Guesty
Get details for a specific property by ID
Get details for a specific reservation by ID
List Guesty properties with optional filtering
List reservations with optional filtering
Output schemas undocumented. LLMs cannot infer what fields list_properties, get_property, check_availability, list_reservations, get_reservation, and create_reservation return. This breaks tool chaining, if create_reservation returns a reservation_id, agents cannot plan to pass it to get_reservation without guessing the field name.
Descriptions are generic and lack actionable context. 'List Guesty properties with optional filtering' (43 chars) does not explain WHEN to call this vs. check_availability, what structure is returned, or how to use the filters parameter. Baseline for A+ tools is 194 chars with explicit WHAT/WHEN/RETURN guidance.
create_reservation lacks confirmation/dry-run pattern. This is an irreversible write operation that creates a booking, agents can make mistakes. No dry-run or confirmation step to prevent accidental reservations.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 64 | 2026-07-28+ | v2 |
Parameter descriptions are minimal. 'filters' is described as 'JSON filtering criteria for properties' but does not explain what fields are filterable, format, or examples. 'fields' parameter lacks guidance on valid field names. LLMs cannot construct valid requests without this detail.
No error handling guidance. If check_availability returns no results, should the agent retry with different dates, call list_properties to verify the property exists, or inform the user? No recovery hints provided.