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

Annotations only say readOnly=false, destructive=true, idempotent=false; the description adds substantive behavior beyond that: no send occurs, line_items overwrites every existing line, tax is retained unless tax_rate/tax_id/remove_tax is passed, and the total is recomputed. It also explains the external consequence that an already-sent client sees the edit behind the same link, which is exactly the context destructiveHint cannot convey.

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-loaded with the editable surface, then moves to behavioral caveats in a compact sequence with zero filler sentences. It loses a point for duplication: "Nothing is sent" is stated twice, and the second instance is wedged into the same sentence as the client-visibility point, blurring the closing line.

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 10-parameter mutation tool with no output schema, the description covers what an agent must know before calling: what is overwritten, what is preserved, how to remove values, that totals are recomputed, and that no delivery happens. Return values are less critical here since the change is described in terms of its resulting state.

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 100%, so per-field meaning is already documented and the baseline is 3. The description earns above baseline by documenting cross-parameter interaction rules the schema cannot express: line_items wiping lines while preserving tax, tax_rate/tax_id alone re-taxing existing lines, and null-as-removal for discount and deposit.

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 (change/update) plus the resource (estimate) and enumerates the exact editable surface: lines, tax, discount, deposit, expiry, notes, PO number. "Nothing is sent" implicitly distinguishes it from send_estimate in the sibling list, so an agent can route correctly without opening 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?

Gives real conditional guidance: "Changes only the fields you pass", line_items replaces all lines, tax_rate or tax_id alone re-taxes current lines, and null on discount/deposit removes them. It never names a sibling tool or states when to prefer create_estimate or send_estimate instead, so it falls short of explicit alternatives, but the when-to-pass-what rules are unusually precise.

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