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

Annotations already declare readOnlyHint=false, destructiveHint=false and openWorldHint=true, so the write/external nature is covered; the description adds workflow facts annotations cannot express — that it returns the same task_id, that polling get_task is required between calls, and that an another_ai reply is kept beside the original in the app rather than replacing it. It does not say what happens if the task is still running or what a failure looks like, so it stops short of full disclosure.

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 precondition and the primary action, and the trailing sentence covers the return contract compactly. The parenthetical "(an option or your own words)" and the mid-sentence another_ai clause make it fairly dense, but no clause 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?

With no output schema, the description correctly compensates by stating the return (same task_id) and the required next step (poll get_task again). It also explains the another_ai side-by-side result, which is the least obvious behavior of the tool. Minor gap: no mention of prerequisites such as upload_file-produced file ids or error conditions when the task is still running.

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 80%, so the baseline is 3; the description earns an extra point by characterizing what `message` should contain (an option or the agent's own words, or a revision request) and by framing `another_ai` as a substitute for the message rather than an addition. It adds nothing about `files` or `doc_ids`, which rely entirely on their schema text.

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 ("send the next message in a task") plus the precondition ("once get_task says it is not running"), and names the sibling tools get_task and list_models that frame it. An agent can distinguish this from start_task or save_reply_to_doc without opening any 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?

Gives the exact gating condition (task not running per get_task), the two legitimate message intents (answer the question or ask for a revision), an explicit alternative mode (pass another_ai instead of a message), and the follow-up action (poll get_task again). When-to-use, when-not-to-use, and what to do next are all present.

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