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

Annotations already declare readOnly=true, destructive=false, openWorld=false, so safety is covered. The description adds the return payload (model and VIP), but says nothing about input requirements such as needing ≥2 groups, group labels aligning with matrix columns, or whether the tool validates the design.

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?

Front-loaded pipe-delimited format with USE WHEN before RETURNS is easy to scan. The ALIASES segment lists six near-synonyms that add little selection value and consume space.

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?

There is no output schema, so the RETURNS line is necessary and useful. However, for a nested-object input with grouping semantics, the description omits the input contract entirely, leaving the agent to guess orientation and label alignment.

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 100%, so baseline 3 applies. The description contributes nothing about data_json structure (samples vs. metabolites orientation) or the relationship between groups and data_json columns, which is the main ambiguity in a two-parameter design tool.

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?

Names a specific method and its discriminative outputs ("supervised discrimination and VIP"), plus a returns clause naming the model and VIP values. It does not explicitly contrast with the nearest sibling, metabolite_pca (unsupervised), so the agent must infer the supervised/unsupervised split from the word "supervised".

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 USE WHEN clause gives clear triggering conditions: supervised discrimination plus VIP extraction for differential metabolites. It offers no exclusions or named alternatives (e.g. 'use metabolite_pca for unsupervised structure'), which keeps it below a 5.

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