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

Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false, so the agent knows this is a safe, non-destructive, idempotent operation. The description adds valuable behavioral context beyond annotations: 'never publishes' (side-effect boundary) and 'never invents metrics' (data integrity guarantee). It also clarifies the input modes ('named recipe or a brief') and output ('compile-ready named-template slides'). This is good additional context that goes beyond the structured annotations.

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 two sentences, front-loaded with the core action and output, followed by two crisp negative constraints. Every word earns its place. It's concise, structured, and immediately scannable.

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?

The tool has 9 parameters, an output schema, and annotations. The description covers the core input modes (recipe or brief) and the key behavioral boundaries (no publish, no invented metrics). It doesn't explain the optional parameters (offer, locale, product, audience, platform, workspaceId) or how they influence the expansion, but the output schema exists and the annotations cover safety. For a complex tool with 9 params, the description is reasonably complete but could benefit from a note on how optional params shape the output. Given the output schema exists, the description doesn't need to explain return values. A 4 is fair.

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 carries the burden of explaining parameters. The description mentions 'named recipe or a brief' which maps to recipeId and brief parameters, but it doesn't explain the other 7 parameters (name, offer, locale, product, audience, platform, workspaceId). The schema itself has enums for recipeId and platform, and types for others, but no descriptions. The description adds some semantic context (recipe vs brief) but doesn't compensate for the 0% coverage of the remaining parameters. Baseline 3 is appropriate because the description adds some meaning but leaves many parameters unexplained.

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 verb ('Expand'), a clear resource ('a named recipe or a brief'), and the output ('compile-ready named-template slides'). It also distinguishes itself from siblings by explicitly saying it 'never publishes' (contrasting with carousels_compile) and 'never invents metrics' (contrasting with generation_plan_brief or similar). This is a clear, specific purpose statement that differentiates it from the sibling tools.

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 implies when to use this tool: when you have a named recipe or a brief and want to expand it into slides. It also gives an exclusion: 'never publishes', which tells the agent not to use this tool when publishing is the goal. However, it doesn't explicitly name the alternative tool for publishing (e.g., carousels_compile) or state when to use a sibling instead. The context signals show siblings like carousels_compile, carousels_clone, carousels_patch_document, which are related, but the description doesn't explicitly route to them. Still, the 'never publishes' clause provides a clear boundary.

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