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

Annotations only say this is a non-read-only, non-destructive, non-open-world write; the description adds the behavior that matters — a closed conversation is reopened, the reply may be delivered by email, private notes are possible, the visitor sees an AI-agent label unless `by_person` is set, and the call is free. Reopening a closed conversation is a state side effect annotations do not disclose.

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 core action, and every subsequent clause adds distinct operational information with no filler. The paragraph is dense and semicolon-chained, so individual constraints (mutual exclusion of point/open_page) can be easy to skim past, but nothing is redundant.

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 7-parameter write tool with nested objects and no output schema, the description covers delivery, side effects, exclusivity constraints, and compliance labeling — enough to call it correctly. It does not say what comes back (message id, delivery status) or what happens on failure, which leaves a small gap with no output schema to fall back on.

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 only 29%, so the description carries most of the load: it explains `point` (highlighting elements in sequence, with the id source), `open_page` (navigating the site), `blocks` (cards with image/title/price/link, images, buttons), `internal`, and the AI-label semantics of `by_person`. Only `conversation_id` and the text length bound go unaddressed, which is a minor gap for such an obvious parameter.

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 — 'Answer the visitor as the business' — and immediately distinguishes a public reply from a private note (`internal: true`) and from rich content (`blocks`). It never names a sibling explicitly (chat_close, ticket_reply, chat_typing), so an agent must infer the boundary from the chat_* family, but the action itself 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 concrete operating rules an agent would otherwise have to guess: replies reach a visitor by email if they left, `point` and `open_page` are mutually exclusive ('one of the two per reply'), and `point` ids must come from page.elements in the chat.message_received event. It stops short of when-to-use vs. alternatives such as `chat_close` or `ticket_reply`.

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