create_conversion_action

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

With no annotations present, the description carries the full burden and delivers exceptionally. It discloses the silent-failure mode (matches-criteria omitted → goal 'creates successfully' but never records), the live-vs-WSDL discrepancy for goal_category (fault 3347), the category-collision fault (5667) with its tested-vs-inferred caveat, and the one-way-door reality (no DeleteConversionGoals; update no-ops; only Paused/ExcludeFromBidding as levers). This is far beyond what is typical and gives the agent honest model of side effects and failure modes.

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 (purpose first, then goal_type bullet list) and logically organized with clear warn-blocks. However, it is a dense wall of text with internal project references ('Task 25e probe', 'Task 25f') and deep fault-code archaeology that a general agent doesn't need verbatim. The information is valuable but could be trimmed to be more efficient without losing meaning.

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 high-complexity tool (7 params, 0% schema coverage, no annotations, no output schema), this is near-complete: it specifies supported goal types, per-type required arguments, category constraints and defaults, the offline-conversion exclusion, the irreversible-creation warning, and documented backout levers. The only residual gaps are the undocumented 'value' parameter and the absence of a note on the expected success/confirmation response, both minor against the depth provided.

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?

At 0% schema coverage the description must compensate, and it does for 5 of 7 params: goal_type (three values, dev-account constraint), url_expression (UrlOperator=Contains), action_expression (ActionOperator=Contains), minimum_duration_seconds (required for Duration), and goal_category (enum values, default, live-verified validity across goal types). The only gap is the 'value' parameter, which receives no explanation at all — a minor omission for an otherwise thorough mapping.

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 a specific action on a specific resource ('Draft a conversion goal') and immediately differentiates itself by spelling out the three supported goal_types and, crucially, what it deliberately does not do (OfflineConversion goals — 'create them in the MS Ads UI instead'). It is clearly distinct from siblings such as get_conversion_actions (read) and set_conversion_action_primary_status (post-creation status control).

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?

Gives explicit when-to-use context: the three goal_type values and their required matching args, plus the lead-gen-meaningful category values. It explicitly states when NOT to use ('OfflineConversion goals are deliberately not created by this tool — create them in the MS Ads UI' with the double-counting rationale) and the one-way-door warning effectively tells the agent to use this only when a permanent, undeletable goal is acceptable. It doesn't name an in-toolset alternative for the offline path, but steers to the external UI, which is clear enough.

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