exp_evaluate

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

Annotations already declare readOnlyHint=true and destructiveHint=false, so safety is covered. The description adds real value beyond that by disclosing that results are deterministic from runs.jsonl and that any user can rerun it, which tells an agent the output is stable and reproducible.

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?

Two tight sentences with no filler, and the behavioral constraint (determinism, rerunnability) is front-loaded. The only weakness is that the opening sentence leads with outputs instead of the action.

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?

An output schema exists, so return values need not be re-explained. However, the unexplained 'name' parameter and the total absence of usage routing leave gaps for a tool sitting in a dense family of exp_* siblings.

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?

The single required parameter 'name' has 0% schema description coverage and is never mentioned in the description. An agent cannot tell whether 'name' is the experiment name, a file, or something else, so the description fails to compensate for the gap.

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 details what the tool returns (metrics per level, CIs, verdicts) rather than stating the action, leaving the verb to be inferred from the name 'exp_evaluate'. It does not distinguish this from siblings like exp_report or exp_compare, which likely also surface experiment results. The purpose lands as vague rather than a clear verb+resource statement.

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?

There is no guidance on when to call this versus exp_report, exp_compare, or exp_runs, nor any prerequisites (e.g., whether an experiment must have been run first). The only contextual hint is 'rerunnable by anyone', which touches idempotence but not selection.

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