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

Annotations already declare readOnly/idempotent/non-destructive, but the description adds substantial behavioral context: the pooled-vs-between-group semantics, the exact error codes raised (nimbus.stats.non_subject_groups, nimbus.stats.insufficient_groups), and the fact that non-qualifying runs return nulls rather than failing the whole request. That is real disclosure beyond the structured safety hints.

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 opening line is well front-loaded and the 'read first' requirement block is deliberately prioritized. It is dense rather than padded, though the final sentence enumerating response labels and the '#470' ticket parenthetical are verbose and partly redundant with the output schema.

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 statistically complex tool with strict preconditions, the description covers the requirement, failure modes, and pooling semantics, and an output schema already exists so return fields need not be re-explained. An agent has everything needed to decide whether to call it and to interpret success vs error.

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%: execution_ids already documents the 1–25 range and the null-reporting/failure behavior, and chance documents the [0,1] range and 1/nClasses default. The description adds nothing new about the parameters themselves (its 'responses label...' sentence concerns outputs, not inputs), so the baseline 3 applies.

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 opens with a specific verb+resource ('Cluster-aware group statistics across executions') and then spells out the exact computation: pooling per-subject accuracies stored by each run's evaluation plan and reporting a between-group mean with a t-interval and one-sample t-test vs chance. No sibling tool (protocol.export_stats, execution.results) covers grouped/cluster statistics, so the agent can identify this tool unambiguously.

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 REQUIREMENT block gives explicit preconditions: executions must have stored per-subject accuracies for >= 3 subjects, session groups within one subject are refused, and single-subject CV fails. It clearly states when the tool applies and when it errors, but never names an alternative tool for the cases it rejects (e.g. what to use for single-subject or session-group stats).

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