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

Annotations already declare destructiveHint=true, idempotentHint=true and openWorldHint=true, yet the description adds genuinely new behavioral facts: the call sends nothing externally, does not resume the interrupted request, and does not authorize a further write. That is exactly the kind of context annotations cannot express. There is mild tension between openWorldHint=true and "sends nothing," but it reads as clarifying that no mutation escapes to the ads platform rather than as a contradiction.

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?

Three short sentences with the operative scope front-loaded and no filler. The wording is dense jargon ("original captured catalog mutation") but every sentence contributes a distinct fact.

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 values need not be described, and annotations cover the safety profile. However, for a six-parameter destructive recovery tool, the description leaves the meanings and prerequisites of work_id, request_key and arguments unexplained and never names the follow-up tool, so an agent cannot reliably construct the call or the recovery sequence from it.

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 50%, and the description adds no meaning to the six parameters at all. Three parameters carry schema descriptions (read_skills, publication_proposal_id, expected_catalog_version) while work_id, request_key and the nested arguments object are undocumented in both places, including how request_key supports the stated idempotency.

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 gives a specific verb and resource ("Save an original captured catalog mutation") scoped to a precise situation ("after interrupted apply"), which is far more informative than the title alone. It does not name the close siblings it must be distinguished from (native.genads_apply_catalog_sync, native.genads_check_catalog_sync, native.genads_catalog_sync_request), so differentiation is left partly implicit.

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 trigger condition (after an interrupted apply) is stated, and the description supplies two negative constraints: it cannot resume the request or grant another write. "An unconfirmed stop can be checked through a separate status read" gestures at an alternative but never names it, so the agent must guess whether that means genads_check_catalog_sync or a status read on the work.

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