Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond annotations (readOnlyHint, idempotentHint, destructiveHint), the description adds substantial behavioral context: pagination mechanics (total, total_na_conta, has_more), the precise semantics of updated_at (record change, not chart viewing), and the stability of id. It also explains that filters are combinable and accent-insensitive. No contradictions with annotations; in fact, it enriches them with actionable details.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured with sections for actions, filters, and notes. It is front-loaded with the core purpose, then details pagination and field semantics. It is somewhat lengthy but justified by the tool's complexity (dual list/get with multiple filters). Every sentence adds value; there is no fluff. The formatting with bullet points enhances readability.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a tool with no output schema, the description provides solid return field details for list (id, nome, link, updated_at, created_at) and summarizes get (full profile with address, phone, panel links). It covers filters, pagination, and relationships to other tools. The only omission is an explicit explanation of the 'account' parameter, but its meaning is weakly implied through 'total_na_conta'. Given the complexity, this is a minor gap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With 0% schema description coverage, the description compensates well: it explains keyword matching (name, nickname, email, phone, CPF, accent-insensitive), updated_since/updated_until (YYYY-MM-DD), order as a filter, page/page_size for pagination, and patient_id/patient_ids for bulk actions. The only gap is the 'account' parameter, which is mentioned indirectly via 'total_na_conta' but never explicitly defined. Overall, most parameters are given clear meaning beyond their schema definitions.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states 'Read patients in WebDiet' and then details two distinct operations: list (paginated, returning specific fields) and get (full profile with address, phone, panel links). It explicitly distinguishes itself from sibling write/delete tools by noting the stable 'id' used across get, update, and prescriptions, and by contrasting with patient_delete's ability to resolve portal links. This leaves no ambiguity about the tool's role.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides concrete usage guidance: it explains when to list (walk pages until has_more is false), how to use updated_since to answer 'who moved in the last N days', and notes that id is stable for use in other tools. It also mentions that only patient_delete resolves portal links, differentiating from alternatives. However, it does not explicitly state 'use this when you need to read patient data instead of X', but the context is strong. Could be more explicit about when to use list vs get, but the details are there.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.