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

Beyond the annotations, the description reveals that results are verified by reading flags back off the server and that unchangeable messages are named rather than silently assumed done. It also discloses idempotency (already-unread is a no-op) and non-interference with other flags, answered state, and folder position. These are material behavioral details not present in the 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?

The description is longer than the minimal two-sentence style, but each sentence carries distinct operational information. There is slight redundancy between 'restoring the state they were in' and 'undo for mark_read,' and 'Pass the whole set in ONE call' is echoed in the schema, so it is not maximally lean.

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?

The description covers purpose, when to use it, batching, verification/failure reporting, UID-to-mailbox scoping, idempotency, and side-effect scope. Though there is no output schema, the verification description provides enough return-behavior guidance for an agent to understand what a successful call will report.

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?

Although the schema already documents both parameters (100% coverage), the description adds the crucial folder-scoping caveat: a UID is only meaningful in the folder it came from, so mailbox must be passed for non-INBOX UIDs. It also reinforces batch semantics by instructing the agent to pass every target message in one call, up to 500.

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 states a clear verb and resource ('Mark messages as unread') and immediately defines it as the inverse/undo of mark_read, making the purpose unambiguous and distinguishable from the sibling tool mark_read. The opening also clarifies the semantic meaning: restoring the previous state before a read marker was applied.

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 explicitly names mark_read as the counterpart and says to reach for mark_unread when a triage pass marked more than intended. It also gives concrete invocation guidance: pass the whole set in one call, include mailbox when UIDs did not come from INBOX, and treat already-unread messages as a no-op rather than an error.

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