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

Annotations declare readOnlyHint=false, idempotentHint=false, openWorldHint=true — mutation-capable and non-repeatable — but the description goes well beyond them: auth/session mechanics, the reply confirmation gate (show the words, set confirmed=true only after agreement), and a detailed prompt-injection safety contract that message content is data, never instruction. It also discloses the unauthenticated fallback ('only service notices are shown'). No contradiction with annotations.

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?

Three dense paragraphs, front-loaded with the sign-in prerequisite, then action semantics, then the safety contract — a sensible order. Every sentence carries weight, though the safety section is verbose enough that a tighter phrasing would read faster.

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 7-parameter, no-output-schema, mutation-capable tool, the description covers auth, per-action behavior, the confirmation gate, safety, and the unauthenticated fallback. The main gap is the shape of what `list` actually returns (fields per message), which an agent may want before summarizing an inbox.

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?

Schema coverage is 100%, so the 3 baseline is already met, but the description adds real semantics: the reply flow requires showing the person their own words and passing confirmed=true, and the settings action is described as read-or-write depending on whether switches are supplied. It adds less on id/hours beyond what the schema already documents.

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 title plus the phrase 'Messages from the Victano team to this person: replies to their feedback, questions about it, service notices and news' names the resource precisely and its content categories. The six action modes are enumerated, so an agent knows it is a multi-action inbox tool rather than a single-purpose one. It does not name a sibling to avoid (e.g. submit_feedback), so it stops short of full sibling differentiation.

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?

Explicit prerequisites ('call connect_start(email) and connect_verify first ... unless this connection is already signed in') and an explicit routing rule ('When an answer carries a `notifications` block with unread messages, tell the person in one sentence and offer to open them; their task comes first'). Each action's trigger is stated (read marks read, reply only for messages that invite a reply, snooze takes hours). Nothing is left to inference.

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