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

Annotations only declare readOnlyHint=false, destructiveHint=false and openWorldHint=true, so the description must carry the real behavioural load, and it does: it discloses that the call blocks (long-poll), that budget_ms bounds the hold and is capped by the host limit, that a timeout returns the same shape with nothing new, and that calls lack a box_key are rejected. The one gap is that it never says the poll consumes/advances events, which is what makes readOnlyHint=false meaningful for a caller deciding whether re-polling is safe.

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?

It is front-loaded with the definition, then the sibling contrast, then auth, then the return shape, with no filler sentences. It is dense and jargon-heavy ('parked', 'mirror', 'suspicious'), which costs a little readability, and the return-field list could be a compact enumeration rather than an inline run-on, but every sentence earns its place.

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?

There is no output schema and this is a blocking, stateful, auth-gated long-poll, yet the description enumerates the returned state keys (state, mirror, parked, game, invite, ended, notices, suspicious), names the enum-like state values, and explains what a timeout yields. Nothing an agent needs in order to call it and interpret the result is missing.

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 all four parameters are already documented in the schema; baseline 3 applies. The description reinforces box_key's provenance (from fyreflock_open's _meta) and ties budget_ms to the timeout behaviour, but adds no new meaning for join or pane beyond what the schema states.

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 opens with a concrete verb+resource: 'Long-polls the assistant conversation for the chat box', and states exactly what resolves the call (a line, a match, an ending, or budget exhaustion). It then distinguishes itself from the closest sibling by name, saying the box uses this while a model relaying for a person uses fyreflock_listen, so an agent can route correctly without opening either 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?

It gives explicit when-to-use guidance ('called only by the box with its box_key'), names the alternative (fyreflock_listen) and the condition selecting it (returns only what is new vs. whole state), and states the exclusion behaviour: without a box_key it is refused with box_only. That is when, when-not, and the failure mode in three sentences.

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