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

Annotations already declare non-read-only, open-world, idempotent, non-destructive, so the description's burden is lighter, yet it exceeds the bar: it names the return payload (conversation.open), explains that the conversation is mediated through the agent rather than an interface, and states how figures/offers are handled with the human's limits enforced. This is behavior the annotations cannot convey.

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 purpose is front-loaded in the first clause and the payload is stated immediately after. It is longer than needed, restating the 'nothing to open' idea three ways ('no app, no chat window, no inbox' / 'there is nothing to open' / 'does not give either human somewhere to go'), but each sentence does carry operational instruction.

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?

Rich annotations plus a named return payload compensate for the absent output schema, and the description covers preconditions, post-call channel mechanics, and the consent/limits model. The only real gap, the source and format of intro_id, keeps it just short of fully self-contained, but for a one-parameter tool this is close to complete.

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% for the single intro_id parameter, so the description must carry meaning. It implies intro_id identifies an introduction where both parties have consented and exchanged first names, which is useful semantics, but it never says where intro_id comes from (e.g., check_in or a prior publish step) or what it looks like.

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 ('Open the direct conversation on an introduction') and adds the precondition under which it applies. It even clarifies what the tool does not do (there is no app or inbox), which sharpens the mental model, though it never explicitly contrasts itself with the near siblings it names (send_message, collect_messages).

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 a clear precondition for use ('once both humans have given the go-ahead and shared their first names') and, unusually, an explicit prohibition ('Never tell them to open an interface and message someone there'). It also routes follow-on traffic to send_message and collect_messages, which functions as alternative guidance, but there is no stated when-not-to-open beyond the consent condition.

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