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

Goes well beyond the annotations (destructiveHint, idempotentHint, readOnlyHint) by detailing the fiscal/accounting impact, the verbatim submission of reason to the tax authority, and the exact semantics of idempotency_key (replay first answer vs. genuinely new cancellations). This is critical behavioral context not visible from structured data.

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 only four compact sentences (or a few short paragraphs), each carrying tactical value: the core action, legal conditions, human-in-the-loop rule, reason constraint, and idempotency behavior. No filler, fluff, or redundant restatements.

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?

Given the tool's complexity and destructive nature, the description covers all non-obvious aspects the agent needs: the legal cancellation window, the requirement to ask the human, the verbatim reason, and high-value idempotency semantics. Since an output schema exists, return values are already documented and don't need describing.

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 compensate. It provides important meaning for two parameters (reason min length and verbatim transfer; idempotency_key retry behavior). Access_key and document_type are left to inference, but document_type's enum is self-explanatory and access_key is predictably the invoice identifier, so the description handles the most nuanced parts.

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 the action clearly ('Cancel an authorized invoice') and grounds it in legal/fiscal/accounting effect within the NFS-e municipality window, which distinguishes it from siblings like correct_invoice or invalidate_numbering.

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?

Explicitly says to ask a human to confirm and never to call it speculatively, giving a clear when-to-use condition. It does not name sibling alternatives for when-not-to (e.g., 'use correct_invoice for adjustments'), so it falls just short of a 5.

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