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

Annotations already declare readOnlyHint/destructiveHint=false, and the description still adds substantial behavior: the branch-checkout vs stacked-view distinction for health, the multi-branch health.branch_id / health.by_branch fallback, rejection of unknown branch ids, and the failure contract where computable:false pairs errors[] (e.g. CIRCULAR_DEPENDENCY) with zero-filled placeholder series. That is exactly the kind of beyond-annotation context this dimension rewards.

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?

Content is dense but load-bearing, and the return-shape sentence is correctly front-loaded ahead of the flag semantics. It is nonetheless one undivided wall of text with heavily nested clauses, and the health/branch material is duplicated nearly verbatim from the schema, so it is not efficiently structured.

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?

With no output schema, the description carries the full burden of describing returns and does so thoroughly: top-level keys, series naming (v, p, vv, ve), health evaluation semantics, and the error/degraded-computation path. An agent can predict the response shape and failure mode without any other source.

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 100% and the schema text is itself verbose (branch_id, health, per_element, granularity all carry their own detailed descriptions), so the schema already does the heavy lifting. The prose largely restates those same semantics rather than adding new ones, which lands at the documented baseline of 3.

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 opening clause gives a concrete verb and resource ('Returns a UID-native snapshot of a model') and enumerates the payload shape (model, items, revision_id/version_number), so the agent knows exactly what comes back. It never distinguishes itself from the sibling layerz_get_model or explains which read tool to pick, leaving that gap against a crowded read-side family.

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 gives strong intra-tool conditions ('variance requires with_values', 'branch_id requires with_values or health', 'granularity ignored when periods are provided'), which is real conditional guidance. It offers no when-to-use versus alternatives, however: nothing tells the agent when to choose layerz_read over layerz_get_model, layerz_diff, or layerz_dependencies.

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