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

Annotations only cover the safety profile (readOnly=false, idempotent=true, destructive=false); the description adds the genuinely non-obvious behavior: optimistic-concurrency failure on stale expected_revision, automatic clearing of answers invalidated by conditions (surfaced in cleared_conditional_answers), and the returned revision/issues/questions. This is exactly the beyond-annotation context that prevents silent data loss.

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?

Four sentences, each carrying distinct payload: the patch guardrail, the revision-conflict contract plus recovery, the conditional-clearing side effect, and the return shape. The most consequential constraint (never guess values) is front-loaded and there is no filler.

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 compensates by naming the return payload (new revision, issues, remaining questions). For a mutation tool with nested-object params, it covers concurrency, side effects, and recovery, so an agent has everything needed to call and retry correctly.

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 the schema already documents the three parameters, including the answers_patch value formats and the null-clears rule. The description still adds decision-level semantics the schema cannot express — only send keys you have answers for, and expected_revision's role as a conflict guard with a defined retry path — which lifts it above the baseline 3.

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?

The description states a specific verb and resource (save answers into a draft application) and the guardrail 'answers the user actually gave' makes the write intent unmistakable, distinguishing it from read siblings like get_application and from create_application. It stops short of explicitly contrasting itself with the other mutating sibling (validate_application), so it lands just under a 5.

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?

It gives concrete usage rules: only include keys you have answers for, never invent values, and on a revision conflict re-read with get_application and retry. That is clear operational guidance, though it frames conditions rather than explicitly stating when to prefer this tool over validate_application or generate_filled_pdf.

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