agent_invoke

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

With no annotations provided, the description carries the full burden of behavioral disclosure, and it delivers extensively. It discloses that invocation is non-blocking, auto-tracks tasks, retries with exponential backoff (default 2 retries), continues running in the background even if the caller detaches, and persists results to task.result/trace. It also details worker-pool assignment, busy-worker fallback, and the semantics of same_worker/auto_fallback — far beyond what annotations would have provided.

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 description is quite long, but it is front-loaded with the core purpose and uses labeled Chinese sections (【后台契约】,【默认worker】,【备路】) to organize complex behavioral rules. Dense parentheticals and mixed-language phrasing reduce readability slightly, but most content earns its place given the tool's 17 parameters and intricate dispatch behavior.

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?

Given no output schema and no annotations, the description does a strong job covering return semantics: it mentions the returned task_id for reuse, task.result for final output, and the panel /api/state endpoint for retrieval. It also covers retries, timeouts, worker selection, and fallback. However, a few parameters remain undocumented (model, workdir, task_id, retry timing), and the description never states what exactly the invocation response contains beyond the task_id, leaving some gaps for such a complex tool.

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 only 41%, so the description must compensate. It does explain many parameters: session_id resumption, max_retries default/disable, timeout_sec for long tasks, plan_mode, same_worker, auto_fallback, fork_on_fail, auto_approval, and capture_trace. However, several parameters — model, workdir, task_id, retry_max_ms, and retry_base_ms — are left unexplained in both the schema and the description, leaving semantic gaps.

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, unambiguous statement: 'Invoke any registered agent by name to run a task.' It further distinguishes itself from the run_* sibling tools by calling itself the 'Generic path over the Agent Registry' and name-driven, which clearly separates it from per-agent tools like run_codex. The reference to agent_list for discovering names adds practical purpose context.

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?

The description explicitly frames when to use this tool versus alternatives: use the generic name-driven path when a new agent (e.g. qwen) lacks a per-agent tool, and consult agent_list to see available names. It also gives clear operational guidance: pass session_id to resume, set timeout_sec for long tasks, use plan_mode for read-only research, and control fallback with same_worker/auto_fallback. This is strong, actionable routing guidance.

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