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

Annotations already declare the read-only, idempotent, closed-world profile, and the description layers on substantive behavior: scheduling semantics for cloud AIs without a wake link ('a normal ChatGPT or claude.ai chat cannot be woken... so tell the owner'), the blocked_by/next_step fields for apps that cannot start work, and the fact that switching apps off happens in the menu bar and account page with no tool for it. That last detail is exactly the kind of negative-space disclosure that prevents wasted tool calls.

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 main verb and resource are front-loaded, and nearly every clause carries non-redundant information about return fields. However, the body is effectively one ~150-word sentence built from nested colons, semicolons and parentheses, which makes the field list (role, app, computer, online, execution_options, profile, blocked_by, next_step) hard to scan and forces re-parsing.

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 carries the return-value burden and does so thoroughly: it enumerates the per-executor fields, the execution_options contents (model, reasoning_effort values, commands), and the blocked_by/next_step/setup_next_step fallbacks. Combined with annotations covering safety, an agent has enough to call the tool and interpret every branch of the response.

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 both parameters (threads_of, reply_session) are already fully documented in the schema, establishing the baseline of 3. The description never mentions either parameter, though its aside that 'chats shows only the newest few' loosely motivates threads_of without giving any calling instruction 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 opening clause gives a concrete verb and resource ('Lists the owner's apps (executors) and saved chats') and defines what an executor is by enumeration (Claude Code, Codex, Cursor, T3 Code, cloud AIs like a Grok Bot). It does not name siblings such as list_tasks or task_status, so the agent must infer the boundary between agent listing and task listing, but the resource 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?

Usage guidance is present but indirect: 'prefer the AI whose profile fits the work' tells the agent how to choose among returned executors, and 'when no app can, setup_next_step says the first thing to fix, for the owner' tells it what to do with a blocked result. There is no explicit statement of when to call this versus list_tasks or task_status, nor any prerequisite (e.g. call before delegate), so the agent must infer the workflow position.

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