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

Annotations already declare the safety profile (readOnlyHint=false, destructiveHint=true, idempotentHint=true, openWorldHint=false), so the description's job is to add context. It adds the tenant-scoping and the all-vs-selected behavior, but says nothing about reversibility, what happens when notificationIds exceeds the 50-item cap, or how the two parameters interact when both are supplied. The 'marks read' framing also sits in mild tension with destructiveHint=true without explaining why the operation is flagged destructive.

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?

A single well-formed sentence that front-loads the verb and resource and places the bulk-mode condition where it is immediately visible. No filler, no restatement of the title.

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?

An output schema and annotations carry much of the load, so the description needn't explain returns or safety. Still, for a destructive default-parameterless mutation it leaves the parameter-interaction rule and the practical path (list_notifications → IDs) unstated, which is a meaningful 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?

Schema description coverage is 0%, so the schema documents no parameter meaning beyond types, defaults and a regex. The description rescues the most important semantics of 'all' (mark every currently unread notification) but never addresses notificationIds' format (ntf_ prefix), the 50-item limit, or which of the two parameters wins when both are provided.

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?

States a specific verb ('Marks ... read') and resource ('tenant-owned notifications') with two explicit modes: selected IDs or all unread. The two operational modes are clear, but it never references the sibling list_notifications that an agent would need to obtain notification IDs.

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?

Usage is only implied: the 'all is true' branch signals a bulk-marks-everything path, and the selected-IDs path implies the caller already has IDs. There is no explicit when-to-use guidance, no warning about the breadth of the all=true path, and no pointer to list_notifications as the way to source IDs.

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