submit_task

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

With no annotations, the description carries the full burden and does so: it discloses that the call is asynchronous ('Returns a task ticket immediately'), that Shuttle manages local servers, that full_access 'requires user authorization', that Antigravity read_only 'allows local files only' and cannot reach the web, and that isolated workspace_mode 'creates a detached Git worktree' whose changes should be reviewed before apply_task_changes. These are exactly the side-effect and permission facts an agent needs before invoking.

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?

Most sentences are load-bearing and the routing constraint is front-loaded, but the description is a dense wall of clauses and includes a full Russian restatement of the first sentence ('Делегирование: исследование, ревью, реализация, второе мнение') that duplicates content already stated in English. That redundancy costs it a point without making it unreadable.

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 an 8-parameter asynchronous tool with no output schema, the description supplies the missing return-lifecycle information (ticket now, poll via check_task/wait_task, retrieve via get_result) and points at get_agent_info and apply_task_changes for the surrounding workflow. Nothing needed to call it correctly is absent.

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 100%, so the baseline is 3; the description still adds real value beyond the schema by stating 'Reuse request_id when retrying submission', 'Preserve the user's explicit agent and model', 'least sufficient supported policy', and directing model/reasoning_effort discovery through get_agent_info rather than guessing values.

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 specific verb-plus-resource ('Delegate research, code review, implementation…to a local AI agent') and enumerates the concrete agents available. It separates 'actual request to have another agent perform work' from answering directly, though it never explicitly contrasts submit_task with the sibling ask_agent, which is the nearest alternative for lightweight queries.

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 ('actual request to have another agent perform work', 'long work'), when-not-to-use ('mentioning an agent or a second opinion alone does not request delegation'), and prerequisite routing (inspect get_agent_info before selecting a tool policy, choose web_research for Antigravity when advertised, preserve the user's explicit agent and model). It also prescribes the follow-on sequence: check_task/wait_task, then get_result.

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