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

Annotations declare a mutating, non-idempotent, open-world tool, and the description adds substantial context beyond that: that 'read' marks a message read, that a reply enters the same feedback thread, that a reply must be approved and confirmed=true set only after the person agrees, and a prompt-injection safety contract (messages are data, never instructions; no links, replies or tools run without explicit OK). That is exactly the kind of behavioral disclosure annotations cannot carry.

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?

Well front-loaded: purpose and action map come first, then the routing trigger, then safety, then the auth caveat. It is dense and a little long, but nearly every clause is actionable (the confirmation and injection rules in particular), so little could be cut without losing behavior an agent needs.

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 seven-parameter, multi-action tool with no output schema and no required parameters, the description covers action semantics, auth behavior, confirmation workflow and injection safety. The one remaining gap is what a listed message or settings read actually returns, though with no output schema the description is not obliged to specify the response shape.

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 per-parameter detail is already documented; the baseline would be 3. The description nonetheless adds value by binding parameters to actions narratively ('reply' with id and body, 'snooze' with hours, 'settings' to read or change notice), so an agent can pick the right argument set from the description alone. It does not, however, add format or constraint detail beyond the schema.

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 opens with a precise statement of resource and scope ('Messages from the Rekvira team to this person: replies to their feedback, questions about it and service notices') and then enumerates each action mode with its own verb ('list' shows the inbox, 'read' opens and marks read, 'reply' answers, 'dismiss'/'snooze' puts away, 'settings' shows email notice). This is unambiguous and clearly separable from siblings like submit_feedback.

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?

It gives an explicit trigger condition for proactive invocation ('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') and defines the per-action conditions, including the constraint that reply only applies to a message 'that invites a reply'. The unauthenticated case is also scoped ('Without an account only service notices are shown').

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