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

Annotations already declare destructiveHint=false, idempotentHint=false, and openWorldHint=false, so the safety profile is covered. The description adds real value beyond that: the operation costs credits, max_credits caps the agreed price, results are fetched asynchronously via list_photos, and a timed-out call may have already started. A '3' would be harsh given that last caveat directly explains the non-idempotent hint — but it never states what happens when max_credits is below the current price, so it earns a 新冠肺炎 4-minus territory; scored 3 for the missing failure semantics.

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 dense but front-loaded: purpose first, then mode semantics, then the price/credit rule, then result retrieval and the retry warning. The first sentence is an overlong run-on covering type=change, and the get_pack parenthetical interrupts the flow. No sentence is wasted, but tighter paragraphing would improve scanning for a 13-parameter tool.

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 a costly, non-idempotent, asynchronous 13-parameter mutation with no output schema, the description covers the critical unknown: how to retrieve the result and how to avoid double-charging after a timeout. Remaining gaps are enum-level behavior of the appearance attributes and what occurs on insufficient max_credits, which are meaningful but secondary.

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 0%, so the description must carry parameter meaning, and it partly does: type mode semantics, text's 300-character limit, max_credits as the user's agreed price sourced from get_shoot/list_photos, and outfit/place as ids from get_pack. However, six appearance parameters (hair_colour, hair_length, makeup, glasses, beard, expression) are collapsed into 'the attributes' with no mention of the 'as_in_photos' default or that these apply only to type=change, and photo_id is never explained. Substantial but incomplete compensation.

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 and resource ('Makes an edited copy of one of the user's photos, paid from their credits') and defines the two modes, 'change' (keeps pose/framing, alters outfit, place, light, attributes, text) and 'zoom_out' (steps back one step). It distinguishes itself from zoom_in implicitly via 'steps back', but never contrasts with siblings like reroll or try_on, so an agent must still infer boundaries. Clear purpose, short of full sibling differentiation.

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 says which type to use for which intent, where to get the price for the required max_credits (get_shoot or list_photos), and where to get outfit/place ids (get_pack of the photo's pack). It also adds an operational rule — check list_photos before retrying an unanswered call. It stops short of explicitly saying when NOT to use this tool in favor of reroll, zoom_in, or upscale.

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