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

Annotations only declare readOnlyHint/openWorldHint/destructiveHint, so the description carries the real behavioral burden and does so: one call per turn, the answer arrives asynchronously as the next user message, the agent must yield its turn, and there is a defined degraded path when no answer comes. It also constrains content ('plain words: never ask about judges, versions or approvals') and enforces a recommended answer per question.

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 purpose first, then constraints, then the turn-handling protocol; nearly every sentence carries a distinct rule with no filler. Some clauses are telegraphic jargon ('One call, one card') that cost a moment of interpretation.

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 human-in-the-loop tool with no output schema, the description covers the full round trip (dispatch, yield turn, resume, fallback) and the question-count and style limits, which is what an agent needs to call it correctly. The gap is the unused 'context' field and the reporting surface of the answer card, which are not described.

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 supply parameter meaning, and it only partially does: it explains the intent behind 'questions'/'recommended' and gives style rules for question text. The 'context' parameter (400-char budget) is never mentioned, and 'options', 'allowFreeText', and the 'id' pattern are left entirely to the schema.

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 gives a specific verb and resource: 'Ask the person up to 3 short clarifying questions before you build or start a non-trivial task,' plus the timing condition. It never names a sibling (e.g., graph_ask_question or agent_run_message), so the agent must infer the distinction from 'the person' rather than being routed explicitly.

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?

Explicit when-to-use ('before you build or start a non-trivial task') and explicit when-not ('Ask only what you cannot find yourself with tools, data or earlier runs'). It also specifies the post-call protocol: 'After the call, end your turn' and the fallback 'If there is no answer... go on with the recommended answers and name them in one line.'

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