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

Despite annotations already flagging destructiveHint=true, the description adds substantial context annotations cannot carry: batched execution, barcode release, partial-failure semantics (blocked rows are reported while others still delete), the fact that HTTP is 200 not 409, posted movement lines surviving as history, and a single-use 5-minute token. It stops short of covering interaction with the unexplained sheet_number parameter, but the behavioral disclosures themselves are unusually complete and do not conflict with the idempotentHint=false / destructiveHint=true profile.

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?

Front-loads the action and the batching limit before the caveats, and most sentences carry unique operational facts. It loses a point for mild redundancy: the blocked-row behavior is described twice ('The first call returns confirm_token and already lists items blocked...' and again in the following sentences), and the block is dense enough to be hard to scan for a destructive operation.

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 destructive bulk mutation with no output schema, this covers the important ground: result shape (deletedCount, blockedCount, results), partial-success behavior, and the confirm flow. The remaining gap is sheet_number semantics and any statement of what the tool returns when the second call is made after token expiry.

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 0%, so the description must do the work. It explains confirm_token thoroughly (must repeat arguments, token expires in 5 minutes, works once, 'A confirm boolean is ignored') and mentions item_ids by name, but sheet_number is never explained — what it selects, whether it is required, or what happens when items span sheets is left entirely to the agent's guess.

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 and resource plus scope: 'Delete up to 100 items on one sheet in one call.' It also explicitly routes small jobs away ('For a single row, delete_item is enough'), which separates it from both delete_item and the bulk_create_items/bulk_update_items siblings.

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 about when to use it versus the single-item alternative, and it names the prerequisite permission ('Requires items.delete'). The two-step invocation is spelled out as a procedure: first call without confirm_token, second call repeating the same arguments plus the token.

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