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

Though annotations already indicate a non-read-only, idempotent write operation, the description goes far beyond them. It discloses that the mapping writes to the integration and rebuilds the feed immediately but still requires a separate re-export to reach the ad platform. It details the exact status response values, the behavior of overridesStepField (including that the rewrite rule permanently overrides a wizard step), the mutual exclusivity of rule with platform_code/constant leading to detachment, and the rejection of rule_enabled alone with no attached rule. All this is invaluable for correct invocation and interpretation of results.

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 long but every sentence earns its place. It is front-loaded with the core purpose, then systematically covers parameters, rules, save/rebuild behavior, response statuses, and edge cases. There is no filler or repetition; the information density is high while remaining logically ordered. For a tool with 9 parameters and many interdependencies, this length is justified and could not be significantly shortened without losing critical guidance.

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 the tool's complexity and the absence of an output schema, the description is exceptionally complete. It explains the return object (including status values and overridesStepField), prerequisites, the distinction between hidden and mappable attributes, the special handling of identifiers with eanFillRate, and the follow-up actions required for re-export. The agent has everything needed to call this correctly without any additional lookups.

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?

With schema description coverage at 0%, the description fully carries the parameter documentation. It explains every parameter explicitly: code (from get_ad_attributes), platform_code (source attribute), constant (fixed value), default_platform_code / default_constant (fallbacks), rule_id (attachment/detachment), rule_enabled (enable/disable), project_id (optional, inferred for single-project), and ad_id (from list_ads). It also clarifies the interplay between rule_id and platform_code/constant and the condition for providing at least one field. This is complete and handles the 0% coverage gap perfectly.

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 precise statement of what the tool does: 'Set where ONE of an ad's output attributes gets its value — the write half of solving the attribute form.' It names the specific verb ('Set'), the target resource ('an ad's output attribute'), and immediately distinguishes it from the read side by referencing get_ad_attributes. This clearly separates it from sibling tools like get_ad_attributes and map_feed_attribute.

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 provides extensive usage guidance: it states when to use the tool (the 'write half' of attribute mapping) and when not to ('If the step already sets the right source, change it there instead of mapping here'). It also prescribes the required prerequisites (get_ad_attributes, list_source_attributes) and even tells the agent to run refresh_ad_category_attributes first for category-specific attributes. It explicitly names alternative actions like repair_ad or run_ad_operation for the re-export step that follows.

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