Uncertainty & Sensitivity

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

Annotations already declare readOnlyHint, idempotentHint, openWorldHint=false and destructiveHint=false, so the safety profile is covered. The description still earns credit by adding non-obvious behavior: 'Pure arithmetic: no I/O and no external calls, rounded to 2 decimals,' the decimal convention for rates, the 1000-100000 iteration constraint for monte_carlo_sensitivity, and that unknown/missing-required-method inputs return an error rather than a value. This is useful context but not exhaustive (e.g. no detail on error shape or output rounding edge cases).

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?

Dense but front-loaded and well-organized: purpose first, then routing, then per-method parameter requirements, then conventions and error behavior. Every sentence carries distinct information (which method, which params, what defaults, what fails), with no filler.

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 an 11-parameter, method-branching tool with an output schema present, the description covers the conditional parameter contract, decimal conventions, iteration bounds, default behavior for unused params, and failure modes. Nothing an agent needs to select a method and supply the right inputs is missing, and return-value explanation is correctly left to the output schema.

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?

Although schema description coverage is 100%, the description goes well beyond the schema by mapping which parameters each method requires (e.g. monte_carlo needs input_distributions; decision_tree needs tree; sensitivity_analysis needs function_name + parameter_name + parameter_range + fixed_parameters), clarifying that only method is required and the rest are method-dependent, and stating the decimal-rate convention and iteration bounds. That conditional cross-parameter logic is not derivable from the flat schema alone.

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 resource ('Uncertainty analysis') and enumerates the four concrete methods (Monte Carlo valuation, sensitivity ranking, decision-tree EVs, one-at-a-time sensitivity), so the agent knows exactly what computation this tool performs. It also distinguishes itself from siblings by directing deterministic point valuations to 'the relevant valuation tool'.

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?

Explicit routing: 'Use to quantify and stress the uncertainty around a point valuation,' contrasted with 'For a single deterministic point value use the relevant valuation tool,' and 'sensitivity_analysis varies one parameter of a core function only.' It also states the when-not condition for each method and notes params of other methods are ignored, leaving nothing to inference.

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