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

Annotations already declare the non-read-only, non-destructive, non-idempotent, open-world profile, and the description goes well beyond them: it queues a pending change that never touches Xero on its own, cannot self-approve, surfaces an approvalUrl, refuses for terminated employees, and leaves no residual Xero-side draft after approval. The only gap is that it doesn't state failure/refusal shape beyond 'refuses plainly' or what a partial/duplicate submission does for a non-idempotent mutation.

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 core behaviour is front-loaded in the first clause and most of the length is load-bearing operational detail rather than filler. It is still a single dense semicolon-chained block where the prerequisite and units instructions could be separated for scanability.

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?

With no output schema, the description usefully names the approvalUrl return field and explains the end-to-end approval lifecycle, which is what an agent most needs. It is undercut slightly by the dangling DATES reference the schema points to, which the description never supplies, so date-format ambiguity is left unresolved despite being flagged.

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 the baseline is 3, but the description adds real meaning on top: employeeId must be an ACTIVE employee id, description is shown to the human approver (max 50), title is AU-only and silently ignored elsewhere, and no units field exists because Xero derives units. However, the startDate/endDate schema entries instruct the agent to 'Follow the DATES instruction in this tool's description' and no such DATES guidance exists in the description, leaving ambiguous numeric dates unaddressed.

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 ('Proposes a DRAFT leave application for one employee/period') plus the exact regional scope (AU/NZ/UK). It also explicitly contrasts itself with create_xero_timesheet and create_xero_pay_run, so an agent can place it among siblings without opening another 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 explicit prerequisites (employeeId from list_xero_payroll_employees, leaveTypeId from list_xero_payroll_leave_types), a negative constraint ('never invented', must be ACTIVE), and an explicit anti-instruction ('Do NOT compute or pass an hours/units figure'). It also names the human-approval step that must follow, leaving little to inference.

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