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

Annotations already declare this a safe, read-only, idempotent, closed-world calculation, so the safety profile is covered. The description adds genuine behavioral context beyond them: the exact penalty schedule applied, and that the result is per-penalty with a charged/avoided/still-to-come status. With no output schema, this return-shape disclosure is the main thing carrying the burden.

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?

Dense but front-loaded: purpose, then the exact computation, then inputs, then return shape, then exclusions. The penalty-formula sentence is long, but every clause carries information the agent cannot get elsewhere. No filler sentences.

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?

A six-parameter calculation tool with no output schema needs to explain both its inputs' roles and its output shape, and this description does both, plus states the excluded regimes. An agent has everything required to call it correctly and interpret the result.

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 description coverage is 100%, so every parameter (including the as_at/filed_on nil-filing distinction and filed_before's effect on the £300 minimum) is already documented in the schema. The description's sentence 'takes the tax year and month, the filing date or none, today's date, the deductions and whether a return had been filed before' mostly restates the schema. Baseline 3 applies.

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+resource ('works out the late filing penalties on one monthly CIS return') and names the governing legislation (Finance Act 2009 Schedule 55), then spells out the actual rate structure. Nothing else in the sibling set computes penalties, so the distinction from cis_deduction, cis_gross_up and cis_tax_month is unambiguous.

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 clear scope boundaries: 'Does not cover deliberate withholding, interest on late payment, or appeals' tells the agent when this tool is the wrong choice. It does not, however, point to any alternative tool for those cases, so routing for out-of-scope questions is left to inference.

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