apply_recommendation

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

With no annotations provided, the description carries the full burden of disclosing behavior. It explicitly states that the tool mutates the account, is not live-verified, faults with 607 on the development account, is gated by an additional flag due to unbounded monetary effect, and that it is a fake. This is comprehensive and goes beyond typical descriptions, covering mutation, failure mode, and risky side effects. No contradictions with annotations since none exist.

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 front-loaded with the action in the first sentence, then provides necessary safety and failure information. Each sentence adds value: the mutation warning, the fake status, and the gating details. It is a bit long but everything included is relevant and not redundant. The structure is logical, moving from function to caveats.

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 mutation tool with no output schema and a single parameter, the description covers key functional aspects: it states the action, the mutation side effect, the gating requirements, the failure behavior, and the fact that it's not live. It doesn't describe what a successful apply would produce, but given that it's marked as fake, that may be moot. It also doesn't mention prerequisites like obtaining the recommendation from list_recommendations, but that is implicit. Overall, quite complete for the tool's context.

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?

There is only one parameter, recommendation_id, and the schema provides its type and required status but no description. The schema coverage is 0%, so the description should compensate, but it doesn't explicitly explain what the ID refers to. However, the context of 'MS recommendation' in the description makes it obvious that the ID is the identifier of the recommendation to apply. While not explicit, the meaning is easily inferred, so a baseline score of 3 is appropriate.

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 states the action clearly: 'Draft applying an MS recommendation (mutates the account — rails apply)'. It identifies the resource (MS recommendation) and the verb (apply). The word 'Draft' is slightly ambiguous—it could imply a non-final action—but the parenthetical clarifies that it actually mutates the account, so the purpose is understandable. It is distinct from siblings like dismiss_recommendation and confirm_and_apply, though it doesn't explicitly name them.

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 does not provide explicit guidance on when to use this tool versus alternatives. It mentions that the tool is 'NOT live-verified' and 'Fakes only', which implies it should not be used in production, but it doesn't direct the agent to a sibling like confirm_and_apply or list_recommendations. The gating details (MS_ADS_ALLOW_APPLY_RECOMMENDATION) are environmental prerequisites, not usage guidance for choosing this tool. Overall, the agent is left to infer when this tool is appropriate.

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