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

Annotations already declare readOnly=false, destructive=false, idempotent=true, but the description adds behavior that annotations cannot express: VIP mail is silently skipped in batch mode, VIP-only actions require confirmation (prompt or vipConfirmed:true), and the result carries a single-use undoId valid for 7 days via undo_view_action. These are non-obvious side effects and gating rules that materially change how an agent calls the tool.

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?

It is front-loaded with the core action and every sentence carries operational content, but the paragraph is dense and mixes several concerns (ID sourcing, sender mode, VIP gating, spam routing, autonomy policy, undo). The inline examples ('archive 2 and 5', 'archive everything from 2') use the verb of a different sibling tool inside a star definition, which reads as copy-paste and slightly muddies the row-number guidance.

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 mutating, nested-object tool with no output schema, the description covers the return contract (counts plus undoId), the undo window and single-use limit, batch size limits, VIP edge cases, and the confirmation policy. An agent has everything needed to invoke it correctly and safely.

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 only 67%, and the description compensates well: it explains where inboxId/messageId come from, defines sender as "every message from this sender still in the inbox" with a 50-newest-per-call cap not stated in the schema, and ties vipConfirmed to the chat-agreement precondition. The row-number workflow is helpful but references row numbers that are not themselves parameters, leaving a small mapping gap for the agent to infer.

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 opening sentence gives a specific verb and resource ("Stars the person's messages"), which is trivially distinguishable from the sibling unstar_messages. The rest of the paragraph makes the scope of the operation (individual messages vs. a whole sender) explicit without needing the schema.

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 spells out two distinct invocation paths with their sources: item rows from show_inbox (with the 'row number' convention), or sender { inboxId, email } after review_unsubscribes. It names exclusions and alternatives explicitly — spam goes through propose_mark_spam then confirm_mark_spam — and states the autonomy rule ("work on your own only on a clear instruction or saved agentRules rule; otherwise suggest and wait").

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