create_character

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

With no annotations, the description carries the full burden and largely meets it: it discloses that this persists a locked-in appearance that governs all future subject_image output, that a bad result locks the whole character into the wrong look, and that force=true is the re-roll escape hatch, plus an instruction to verify the returned reference image. It is silent on the non-appearance generation knobs (seed, guidance_scale, steps, size) and on collision behavior when a name already exists.

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?

Well front-loaded – the one-line purpose precedes the rationale and the field guidance – and structured into clearly labeled sub-sections that an agent can scan. It is on the long side, and the illustrative anecdote about the robe-to-armor swap could be trimmed, but essentially every sentence carries usable instruction.

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?

An output schema exists, so return-value explanation is not owed, and the description correctly just tells the agent to inspect the returned reference image. Combined with the purpose, usage and appearance guidance, an agent can call this correctly; the gaps are the undocumented generation parameters and name-collision behavior rather than anything blocking.

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 0%, so the description must compensate. It does so richly for the two most important parameters: `appearance` is decomposed into identity vs. default clothing with a concrete list of what must be pinned down, and `force` is defined as the re-create/override switch. The other five parameters (seed, width, height, guidance_scale, num_inference_steps) receive no semantics at all, leaving a substantial documentation hole.

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 sentence names a specific verb and resource ('给一个人物定妆' – generate and persist a character reference) and immediately contrasts it with generate_image and subject_image, so the agent knows it is not a one-off image generator. It does not, however, differentiate itself from the parallel siblings create_actor / create_animal / create_object, which is the one ambiguity left.

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?

Explicit about when to use it (whenever a character must stay visually consistent across many images), what to use instead for one-offs (generate_image), and what NOT to put in this call ('不要写场景、动作、表情' – leave those to subject_image's scene). It even gives a worked example of swapping clothing via scene without re-creating the character, which is exactly the when/when-not guidance an agent needs.

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