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

Annotations already declare destructiveHint=true and openWorldHint=true, so the safety bar is lower. The description adds real behavioral context beyond them: even split is the default, holdout arms never produce leads so they measure exposure only and not lift, and extra arms each require the same traffic again. It does not disclose what starting the experiment does to existing traffic on the control widget or whether it is reversible.

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 with purpose, then prerequisites, then split semantics — a sensible ordering. It is a single dense block of prose that could be broken up for scanability, but nearly every clause carries information an agent needs.

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?

Covers prerequisites, split math and result-reading follow-up for a destructive, open-world mutation. The main gap is the return value: with no output schema, the description never says whether the call returns an experiment id needed by read_experiment_results or stop_experiment.

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 only 20%, but the description carries the load for every functional parameter: widgetId as the control, variantWidgetIds as 1-3 ordered live variants, trafficPcts as whole-number percentages with the control first and ordering tied to the variant list, and holdoutPct's range, default and complement-to-100 relationship. Only `project` is left to the schema, which documents it.

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?

Opens with a precise verb and resource: 'Split traffic between a control widget and 1 to 3 variants.' An agent can distinguish it immediately from read_experiment_results and stop_experiment without opening any schema.

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?

States hard preconditions — variants must already exist as live widgets of the same kind and project, created with create_widget and published with request_widget_publish — and routes to read_experiment_results for outcomes. It also flags a practical constraint ('on a small site one variant at a time'). It stops short of explicit when-not-to-use conditions or naming stop_experiment.

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