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

Annotations already flag destructiveHint and idempotentHint, but the description adds non-derivable behavior: activation goes live immediately with no separate publish step, and a held change set is only applied once complete. The admin-only requirement and consent gate are extra operational context. The phrasing around held changes ('lands when it is whole') is vague enough to leave the reader guessing about partial states.

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?

Four short sentences, each carrying distinct information: scope, effect, publication semantics, consent, permission. The terse, elliptical register ('Only when the person says so') is a style choice rather than padding, and nothing is repeated from the schema or annotations.

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 two-parameter mutation with no output schema and full annotation coverage, the description covers scope, immediacy, consent, and privilege. What's missing is recoverability — nothing says whether the previous theme can be restored via `restore_theme_revision` or `theme_history`, which matters for a destructiveHint tool.

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 coverage is 50%: `site` is documented in the schema as a name from `list_sites`, which the description echoes rather than extends. The `theme` parameter has no schema description at all and the description never clarifies whether it takes a name, ID, or handle — the one place prose was most needed.

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 concrete verb and resource: making a given theme the one a site 'wears' live, scoped to the site named in the `site` parameter. The metaphor 'wears' is slightly indirect for what is really theme activation, but combined with the title and the reference to `list_themes` it is unambiguous enough to distinguish this from `preview_theme`, `build_theme`, or `update_theme`.

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?

'Only when the person says so' is an explicit consent gate, and 'Admins only' names the required privilege — both are real usage conditions rather than restatements. It stops short of naming sibling alternatives (e.g., use `preview_theme` to inspect first, `restore_theme_revision` to roll back), so it stops at 4.

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